/i "61 N No. 16. T. P. r.-G6. U. S. DEPARTMENT OF AGRICULTURE. DIVISION OF VEGETABLE PHYSIOLOGY AND PATHOLOGY. \. 3). It was also found that a uredo form found common on Kentucky blue grass would not infect wheat nor oats.' The conclusion was then expressed that " these experi- ments seem to show that the rusts of various cereals are probably physiological species." At the time this work was done the writer was not aware that Bolley had made some experiments of the same kind. BoUey, however, reached no conclusion (10, p. 2G3). On March 1, 1894, the writer began work for this Department, con- sequently the inoculation experiments just described were dropped and were not resumed for nearly two years. In the meantime Eriks- son, who in 1890 began experiments in the same line, published numer- ous articles^ giving many results corresponding with those above mentioned and with others which have not yet been published. In some instances, however, the results difier. Specialized forms of cereal rusts. — The species hitherto known all over northern Europe under the one name, Puccinia rubigo-vera (DC.) Wint., have been separated by Eriksson and Hennings into two dis- tinct species, i. e., F. (flumarnm (Schmidt) Eriks. and Heun. (the "yel- low rust)," and P. disjiersa Eriks. and Henn. (the "brown rust") (30, pp. 197, 198, 257, 258). Three rusts, therefore, attack wheat in Sweden, the yellow rust bemg, it is said, the most common and destructive. So far as known this yellow rust does not occur in America, but the brown ' Swell ideas had already been expressed by the writer in Science, Vol. XIV, p. 63. i^On page 4 of the Kansas Agr. Expt. Sta. Bull. No. 46 it is stated that Uredo graininis of oats would not infect orchard grass, but subsequent investigations have shown that this grass is a common host for that rust. 5 Eriksson, Jakob, Die Getreideroste, prepared in conjunction with Ernst Henniugs and published originally in Swedish, comprises the details of Eriksson's investiga- tions from 1890 to 1893. It did not reach America until the summer of 1896. The principal results were given in Die HanptresuUate einer neuen Untersnchung ilber die Getreideroste (Zeitschr. f. Ptiauzenkrank., 1894, Vol. IV, pp. 66-73, 140-142, 197-208). The first part of this article appeared one month before the results obtained by Hitchcock and the writer were published, but was not seen by them for some time after. Ueber die Speciahsirung dcs Parasitismus bei den Getr eider ostpilzen {Ber. d. Deutsch. Bot. Gp- , 1894, Vol. XII, pp. 292-331) gave the results of his further experi- ments for 1894, and also references to the work of Hitchcock and the writer. Sub- eeciuent to 1894 Eriksson's researches have been reported in various papers. 10 CEREAL RUSTS OF THE UNITED STATES. rust is probably the same as our P. rubigo-vera. These authors also make the rust heretofore called P. rubigo-vera simplex, occurring on barley, a distinct species — that is, P. simplex (Kcke.) Eriks. and Henu. (30, p. 260). The writer has no reliable information as to the occur- rence of this species on barley in this country, but what is possibly the same rust has been found occasionally on native species of Hordeum. In his inoculation experiments Eriksson also has observed that forms of a certain rust species taken from certain hosts would not infect certain other hosts that are attacked by this same si:)ecies. This may be true even where the hosts are of the same genus. On the other hand, infections will sometimes take place from the host species of one genus to the species of another genus. To such distinct forms of the same rust species he has given the name forma specialis. The form name is taken from the generic name of one of the hosts to which the form is restricted, preceded by the abbreviation f. sp. As there seems to be no good reason why the entire name should not be written as an ordinary trinomial, designating the variety, it will be so written in this bulletin. These forms are, as a rule, as good varie- ties as those established on morphological grounds by phanerogamic botanists. Eriksson's researches have led him to conclude that so far as known there are at present ten forms of these cereal rusts in Sweden, that is, three black rusts, Puce in ia graminis secalis on rj^e and barley, P. gram- inis tritici on wheat, and P. graminis avenw on oats; three yellow rusts, P. glumarum secalis on rye, P. glumarum tritici on wheat, and P. glii- mariim hordei on barley; two brown rusts, P. dispersa secalis on rye and P. dispersa tritici on wheat; one dwarf rust, P. simplex on barley; and one crown rust, P. coronifera on oats. The writer's inoculation experiments in 1896 and 1897, several thousmid in number, which will be described in detail further on, showed that there are now at least six distinct forms of cereal rusts in the United States, and probably a seventh, namely, two orange leaf rusts, P. rubigo-vera tritici on wheat and P. rubigo-vera secalis on rye; one crown rust, P. coronata on oats; three black stem rusts, P. graminis tritici on wheat and barley, P. gra- minis secalis on rye, and P. graminis avenw on oats; and one maize rust, P. sorghi on maize. In comparing the Swedish and the American rusts, it should be observed that Eriksson uses for the crown rust the name established by Klebahn (42, pp. 134-136), that is, P. coronifera. He justifies this ui)on the basis of his own (28) and upon Klebahn's experiments, which showed that in northern Europe P. coronata Corda is a collective spe- cies, of which one section has Bhamnus cathartica for its ajcidial host and occurs on oats and other grasses, and another section has li.fran- gula for its .Tcidial host, but does not occur on oats. The former is called P. coronifera and the latter is called by the old name P. coronata. Each of these newly made species has also its specialized forms or varieties, there being eleven in all. As the writer has not yet deter- INTRODUCTIOX. 11 niiued the a^cidial host of the crown rust of oats for this country,' it is deemed best to retain, for the present, tlie okl name P. voronata; and while Eriksson and lleniiings's division of P. ruhigovem is undoubtedly correct, it seems, nevertheless, preferable to retain this well-known name'^ until it is positively determined whether our species is equiva- lent to P. dispersa or is a new one. It has already been stated that P. ylumarum and probably P. simplex do not occur in this country, and of course P. sorghi does not occur in Sweden. Common nomenclature of grain rusts. — The common nomenclature of these rust forms is a complex question. The names above given were decided upon, after much consideration, as being the most accurate and descriptive that could be used. In the case of 7*. graminis and P. ruhigo-vera the names have been given in accordance with the most common and striking characteristics of the species. The former, when it occurs in abundance, exists principally upon the stems and in the black stage, while the latter is most strikingly exhibited on the leaves in the orange-colored uredo stage. The crown rust, although also an orange leaf rust, can not be spoken of as such, as it would thus be con- fused with the orange leaf rust of wheat, which is a different species. Of course, there need be no confusion in the case of the name of the maize rust, it being the only one on maize/^ The Departmenfs irork on grain rusts. — AYith a view to thoroughly investigating the cereal rust question, this Division in the spring of 1892 sent to crop reporters in the principal wheat-growing States a cir- cular requesting information as to the distribution and abundance of the different rusts, the damage caused by them, etc. Owing to the press of other work, however, nothing further was done in this line, except the spraying experiments in 1892 and 1893, until the writer took charge of the work in the spring of 1894. A circular similar to the above was then prepared and sent to growers in all States where wheat and oats are grown to any extent, the information requested to be based on observations made during the season of 1894. The following table gives a combined summary of the answers received to the two sets 1 Siuce the above was written a series of inoculation experiments have shown rather conclusively that Ehamnus lanceolaia acts as an tecidial host of the cro>vn rust in this country. ^The writer has recently had au opportunity to make personal observations on the rusts of wheat in northern Europe — that is, from England to Sweden and south- ward to southern Germany — and not only found these two to be very distinct, but also found that P. glumaruvi was far more common and injurious everywhere during the season of 1898. It seems strange that the two species should have been confounded for so long a time. However, there appear to be very good reasons for still retaining the name P. ruiigo-vera for the species called P. dispersa by Eriksson and Heuniugs, as there is little doubt that it is the same as the one so well known under the former name in all other countries where rusts have been studied. *The fact that Eriksson called P. dispersa brown rust — and, judging from his speci- mens, he seems correct— is the only reason for doubting that P. ruMcjo-vera of this coun- try is equivalent to it, as our species is certainly orange color in the uredo stage. The uredo of the black stem rust would be more properly called brown. 12 CEREAL RUSTS OF THE UNITED STATES. of circulars. Kortli Carolina and Georgia, though not important as wheat States, are included in the table to show the conditions respecting rust in such latitudes. Table 1. — Summary of reports on the cereal rusts received in 1892 and 1894 from twenty-three States. States from ■which. Reports ou wheat- Per cent of damage.* Part of plant chiefly affect ed by rust. Reports on oats — Part of plant chiefly affected by rust. reports were re- ceived. Injured by rust. Free from rust. rr. s = , > CG 1-5 o Injured by rust. Free from rust. as 1 X > 9 7 1 6 27 27 47 32 34 43 3 18 5 27 38 7 12 30 31 4 43 61 7 4 73 54 87 35 90 22 2 2 12 23 48 26 43 77 105 10 9 4 15 23 ( 2) 30 (26) 17 ( 2) 31 ( 7) 20 (35) 19 (35) 24 (44) 25 (30) 26 (49) 40 (52) ■ li 1 1 4 4 2 1 "2 4 10 2 1 3 6 27 53 4 1 63 29 72 33 69 40 12 11 1 2 13 9 3 3 11 8 10 4 2 4 1 4 4 3 5 2 1 5 7 2 2 Pennsylvania 2 2 North Carolina 2 Ohio ...... 1 7 6 4 i 28 26 44 21 22 19 7 11 12 16 23 5 10 32 50 6 3 3 10 ...... ...... 1 ■.■.'.".■.; ...... ...... ...... 1 1 3 !Michicaii i 4 ; 13 1 2 14 1 8 1 5 16 2 2 4 Tennessee 2 1 25 (10) 4 "2 1 1 2 2 "'5' 1 4 30 40 21 20 67 70 3 2 iMinnesota^ 2 2 4 46 (16) 26 (52) 11 (10) 21 (10) 28 (37) 25 (38) 12 ( 1) 2 ATisHoiiri ........ 1 8 3 North Dakota South Dakota Nehraska ....... 1 1 4 3 2 4 Tvansas .... ...... 5 Texas'^ 2 1 3 1 46 ( 3) 1 1 2 3 2 States from which reports were re- ceived. Varieties of wheat free from rust. ^ New York Pennsylvania. Clawson (3), bearded varieties (2). Pultz {13), Red Lancaster (6) .. . Varieties of wheat liable to rust. Clawson (3), white varieties (2). Late varieties (5), bald varieties (3). ruitz(2) Ford (1), May (1) Maryland^ ' Fulcaster (2) North Carolina* .. . Fulcaster (3), bearded varieties I (!>• Ohio ! Early varieties (12), Fultz (9) ..I Late varieties (8), Clawson (7) . Michitran ' Clawson ( 12), Fultz (7) Clawson (14) , Mediterranean (4) Indiana Fultz (27), early varieties (14) .., Late varieties (20), bald varie- 1 ties (9). Wi^iconsin 1 Fultz (3), hard varieties (3) Soft varieties (2), Fultz (2) niinois ! Fultz(19), bearded varieties (10). Fultz (10), bald varieties (8) Kentuckv Tennessee 2 Georgia' . .. Minnesota 2. Iowa Missouri North Dakota. South Dakota. Nebraska Kansas Texas' Oregon ' . . California Fultz (10), bearded varieties (9). Rice wheat (2), Missouri Red (1) Blue Stem (3), Dallas (2) Saskatchewan Fife (1), Blue Stem (1). Blue Stem (7), early varieties (3) . Fultz (32), Mediterranean (21).. Blue Stem (4), hard varieties (3). Blue Stem (14), hard varieties (4) Velvet Chafl'(lO), Blue Stem (6). Turkey (32). Russian Bearded (ID- Mediterranean (4),]Sicaraguan (1). Sonora (2), Odessa (2) Bald varieties (8). white varie- ties (6). Bald varieties (2), soft varieties (1). Walker (1). Moore (1) Soft varieties (1) Bald varieties (3), soft varieties (2). Walker (7), Fultz (7) Scotch Fife (3), White Russian (2). Soft varieties (5), Lost^ation (3) Bearded varieties (6), Grass (5) . Soft varieties (15), May (8) Wheat most affected. Early. Late. Golden Chaff (1), bald varieties (!)• RedChSffd) White Club (2), White Austra- lian (2). 32 45 3 5 64 61 101 53 75 49 14 2 32 80 20 23 63 62 1 The figures in parentheses show the number of replies from which the average per cent was calculated. 'Thereportfrom this State is for 1894 only. . 3 The two varieties mentioned in each case were these given the greatest number ol times in tne replies. The figures in parentheses indicate the number of replies in which the variety was namea. RUSTS IN DIFFERENT STATES. 13 The writer has visited, in some cases at ditfereut tiraes, all the im- portaut wheat-growing States except Pennsylvania, Washington, Oregon, and California. The most extended tour was made in 1894, when all the Avheat-growing States of the Great Plains and of the Mississippi Valley were visited. However, on account of the drought which prevailed during that season the conditions were in some respects most unfavorable for a study of the rusts. To gain a very accurate knowledge of conditions in the field a personal inspection of cereal districts for three to five successive years would be necessary; but, nevertheless, the observations just referred to, together with the replies from correspondents, have furnished valuable information and have made the situation much clearer than before. GENERAL OBSERVATIONS ON CEREAL RUSTS IN DIFFERENT STATES. New YorTc. — Kust does not seem to be as prevalent in this State as in some others. Oats is often injured, especially in Erie and Wyoming counties. Considerable spring wheat is raised, and it is always more rusted than fall wheat, as are also late-sown wheat and oats. There is considerable stem rust, but to what extent it prevails over the State is not definitely known. Some counties in which rust is quite preva- lent are Franklin, St. Lawrence, Cayuga, Schuyler, Wayne, Clinton, Greene, and Cattaraugus. Pennsylrama. — The reports indicate tbat wheat is not extensively damaged by rust, but that oats is very commonly injured. It is quite probable, however, that much of the damage attributed to rust is due to "oat blight," a supposed bacterial disease occurring in this State, which gives the crop a yellowish red appearance. Fertilizers are sup- posed to materially aid the grain to escape rust by causing rai)id growth and early maturity. The rust is always worst in the lowlands. Among the counties in which rust is especially prevalent are Columbia, Adams, Berks, Franklin, Northampton, Schuylkill, Forest, Sullivan, and Lackawanna. The standard varieties of wheat grown in this State are Fultz, Lancaster, and Fulcaster. Maryland. — It is claimed that the application of fertilizers, especially phosphates, has materially decreased the losses from rust in this State in recent years. In the experiments conducted by the Department at Garrett Park during three seasons there was an abundance of orange leaf rust on wheat and crown rust on oats, but they did not appear to do any particular damage. Black stem rust was not observed. As in Pennsylvania, oat blight produces much of the injury commonly ascribed to rust. Much oat rust has been reported from Cecil County. North Carolina, — Wheat is quite commonly injured by rust, but it is not grown extensively in this State. The orange leaf rust seems to be the prevailing species, though as yet it is uncertain whether it is the only one present in cases of severe injury. One peculiar fact of interest is that early wheat, such as Early May, is often more injured than late 14 CEREAL RUSTS OF THE UNITED STATES. varieties, wliich is just tlie reverse of tlie rule. A probable reason for this is that early wbeat is sometimes weakened by frosts or freezes and is thus rendered more liable to rust. In the Atlantic coast States the wild blackberry rust {Cccoma niteris) is extremely abundant, especially in the edges of clearings near fields of wheat— a fact which has given rise to the erroneous opinion in this State that this rust has some ontogenetic connection with the cereal rusts (53, p. 16). Three counties in which rust is quite prevalent are Eandolph, Alexander, and Moore. Ohio.— The black stem rust is occasionally very prevalent in this State. There is much more injury by rust to oats than to wheat. On account of early ripening, Fultz wheat is less liable to rust than other varieties. Here, also, fertilizers cause rapid growth and early maturity, and act to some extent as preventives of rust. The counties in which rust is reported to be especially injurious are Clark, Stark, Kichland, Fayette, Hancock, Knox, Logan, Guernsey, Crawford, Scioto, Defiance, Greene, Medina, Mercer, Meigs, Miami, and Harrison. Michigan.— In this State wheat is often badly damaged by rust, but, as in all the North Central States, oats suffers still more. Wheat is much more liable to rust injury after winterkilling, hence late varieties suffer more than the early varieties. The varieties of wheat most gen- erally grown are Clawsou and Fultz. Rust rarely occurs in the sandy districts. It is apparently most abundant in Antrim, Jackson, Wayne, Charlevoix, Cheboygan, Lapeer, Benzie, Alcona, Leelanau, Clinton, Shiawassee, Livingston, Washtenaw, Osceola, Midland, Saginaw, and Ingham counties. Indiana.— V^heat is much injured and oats is commonly damaged in this State by rust. Late varieties are invariably much more liable to rust than the early varieties. Fultz wheat escapes rust, not because it is rust resistant, but because it ripens early. In some instances the growing of oats has been abandoned on account of rust. Grain is said to be free from rust in the warm, sandy soils. Some claim that it is very severe only once in two to five years. Black stem rust is the most injurious species occurring in this State. The counties in which the greatest amount of rust is found are Wayne, Yanderburg, Jen- nings, Pike, Sullivan, Greene, Ohio, Union, Lawrence, Posey, Gibson, Clinton, Ilipley, Bartholomew, Huntington, Adams, Morgan, Wells, Clark, Delaware, Miami, Starke, and Johnson. T^l-sconsiH.— According to nearly half the reports received from this State, wheat is commonly damaged by rust. As in some other States, it is believed by many that an application of salt, lime, or ashes hastens ripening and makes stifler and cleaner straw. It is claimed by some that White Kussiau oats possesses a considerable degree of rust resistance. Black stem rust constitutes a good proportion of the cereal rusts. Some t»f the counties in which there is much rust are Marathon, Pepin, Waukesha, Richland, Ozaukee, Brown, Waupaca, Adams, Marquette, Dunn, Waushara, Polk, Sauk, Outagamie, Juneau, Bufl'alo, Winnebago, Saint Croix, Jackson, and Sheboygan. RUSTS IN DIFFERENT STATES. 15 Illinois. — Oats is commonly and often seriously damaged by rust. The damage to wheat is much less, although it, too, often suffers severely. In 1890 and 1S91 rust was unusually severe in some locali- ties. It is always worst iu fields of grain that have winterkilled. The damage is particularly severe once in three to five years. Wheat seems to suffer more in the extreme northern and southern portions of the State than in the central part— due in the southern part probably to the latitude and climate, and in the northern part to the large propor- tion of spring wheat grown there, but spring wheat is rapidly losing- prestige and is not grown so extensively as formerly. Fultz wheat is a popular variety throughout the State. In some localities tile drainage is said to reduce the amount of rust. Some believe that although bearded varieties are less liable to rust on prairie lands, they are more liable to it on timber lauds, but on what this belief is based the writer does not know. Some of the counties in which much rust seems to occur are Hardin, Morgan, IMassac, Williamson, Whiteside, Edgar, Dekalb, Grundy, Cass, Menard, Cook, Boone, Lee, Carroll, Woodford, Stark, Pope, Lawrence, Stephenson, Pulaski, Johnson, and Jersey. Kentucky. — The reports of correspondents, as well as the writer's observations, indicate that in this State wheat is very commonly and severely damaged by rust. In other States the comi)lete destruction of the crop occurs occasionally, but in this State this is a common occurrence, lilack stem rust is very common and, as it ai)pears later than the orange leaf rust, the introduction of early- maturing varieties is a matter of much importance. Some make a practice of harvesting the grain while in the "dough'' stage in order to escape the rust. Oats is also extensively injured. Some of the counties where the grain is most affected are Trigg, Harrison, Metcalfe, Simpson, Jackson, Caldwell, Henderson, Nelson, Allen, Garrard, Grayson, Adair, Magoffin, Crit- tenden, Fayette, Casey, Calloway, Webster, Christian, Livingston, Letcher, Grant, Graves, McLean, Ballard, Muhlenberg, Marshall, Martin, Lawrence, Estill, McCrackeu, Kenton, Montgomery, Knox, Clay, Monroe, Boone, and Ohio. Tennessee. — The information concerning rust in this State is still quite meager, though the writer has made two trips of observation to different localities. However, the indications are that wheat is rather commonly injured by rust. The stem rust is very severe some years, but seems to be less common than in Kentucky. Little injury is done to oats. Kice wheat is considered by some to be a rust-resistant variety. Fultz and Fulcaster are favorite varieties in some localities. Stem rust becomes abundant from May 20 to June 1. Greene County seems to have considerable rust. Georgia. — Yery little wheat is raised in Georgia, but usually what is grown is badly injured by rust. Leaf rust is the species most prevalent, hut frequently the grain is not severely injured until the stem rust appears. In this State, as in all the Gulf States, oats is seldom injured by rust. The variety of oats most commonly grown is the Texas Eust 16 CEREAL RUSTS OF THE UNITED STATES. Proof. Scarcely a county in the State is free from rust, but it appears to be most abundant in Lincoln, Jones, Milton, Catoosa, Hart, Kau- dolph, Bartow, Henry, Marion, and Sumter counties. Minnesota. — Xot much information is at hand concerning rust in this State, but it is known tliat in certain years both wheat and oats are greatly injured. The Fifes and the Blue Stem are the standard varie- ties of wheat generally grown. White Russian oats is said by some to be rust resistant. Reports of rust have been received especially from Stearns, Fillmore, Renville, Ottertail, and Watonwan counties. loica. — So great has been the damage from rust in this State that in some parts wheat growing has been gradually abandoned principally from this cause. In Buena Vista County it is reported that there has been an average annual loss of 10 per cent of the crop during the past twenty-two years. This severity of the rusts is doubtless due to the practice of spring sowing, so common in the State. The great desid- eratum for Iowa in the way of cereals is a good, hardy wheat like the hard red wheats from south central Russia, as the Ghirkas, Crimean, Tx, etc. The amount of oats sown in the State as compared with wheat is becoming quite large, but rust often destroys this crop also. It is always the black stem rust that is particularly injurious. Rust is pretty evenly distributed over the State, but it is probably particularly abundant in Fremont, Buena Vista, Des Moines, Dallas, Clay, Butler, Crawford, Cerro Gordo, Appanoose, Buchanan, Cedar, Cherokee, Greene, Franklin, Emmet, Allamakee, and Plymouth counties. Missouri. — There is considerable rust in this State, but not much is known concerning the relative proi^ortions of leaf rust and stem rust in cases of great injury. Mediterranean wheat is considered to be rather rust resistant. Oats is often much injured. Rust seems espe- cially common in Christian, Crawford, Bollinger, Carroll, Atchison, Iron, Oregon, Dent, Dekalb, Franklin, Carter, Hickory, Livingston, Mercer, Sullivan, Daviess, Madison, Grundy, Kodaway, Gasconade, Linn, Jeflersou, and Perry counties. Xorth Dakota. — Wheat does not seem to be commonly damaged in this State, although leaf rust is always present and is sometimes abundant. Oats, however, is often injured. White Russian oats is thought to be somewhat rust resistant. The Fifes and Blue Stem (Vel- vet Bald Blue Stem) are almost the only varieties of wheat known by the majority of the farmers. By far the greater jiroportion of the cereals of the State is, of course, grown in the extreme eastern part, in or near the fertile James River Valley, and the rusts naturally occur in greatest abundance in that region. Nevertheless the orange leaf rust and occasionally the stem rust also are found far westward. Much rust is reported, esj)ecially from Nelson, Grand Forks, Cass, Richland, Steele, McHenry, and Kidder counties. South Dakota. — In this State, as in North Dakota, it is, as a rule, too dry, cool, and breezy for rusts to do much damage, but nevertheless RUSTS TX DIFFERENT STATES. 17 tliey are occasionally ([iiite injurious in places. The leaf rust is the species most common. The Blue Stem and various Fifes are the varie- ties of wheat generally grown. Oats, especially when planted late, is often batlly rusted. Naturally rust is most abnndantii! the southeast- ern counties. Hust seems to be quite severe, at least occasionally, in Kingsbury, Clay, Deuel, h^uilk, I'rookings, Hanson. Minnehaha, Union, Douglas and Davison counties. Xebraska. — Usually wheat is not batlly damaged by rust in this State, although sometimes the injury is quite serious. The cereals in large portions of the State were badly injured in 1891, the black stem rust being abundant that year. In the sandy districts there is usually but little trouble with rust. As in Iowa, spring sowing is common, and is doubtless res[)ousible for much of the injury. Fortunately, however, fall sowing is coming into vogue. As is usually the case, rust often accompanies winterkilling and poor drainage. In the eastern portion of the State oats is often severely damaged. The counties from which much rust is rei)orted are Douglas, York, Loup, Dodge, Nuckolls, Blaine, Gage, Custer, Frontier, Nance, Perkins, Hamilton, Greeley, Howard, Adams, Wayne, Cuming, Boone, Seward, and Saline. Kansas. — In the eastern half of the State the wheat crop is (|uite often injured by rust, but in the western half the damage is usually insignificant. Notwithstandang its com}»arative immunity, the writer once saw a held of White Michigan wheat in Cloud County totally destroyed by rust, the i)robable yield of the field having been esti- mated a few days prior to the attack at 2o bushels per acre. Turkey wheat is quite resistant to rust and is a general favorite in the State, particularly in the western part, and yet even in that section the writer has seen it severely attacked by the stem rust. Oats is often greatly damaged by stem rust in the eastern portion and is sometimes a total loss. The crop suffered greatly in 1893. Occasionally fields of late-sown oats are entirely destroyed before harvest. The counties in which rust seems to occur most frequently are Kice, Montgomery, Johnson, Stevens, Washington, Linn, Riley, OvSage, Jewell, Keno, Marshall, Barton, Coffey, Bourbon, Ottawa, Shawnee, Decatur, and, Miami. Texas. — Reports received from the State during the dry season of 1894 indicate that grain is not usually attacked by rust. However, but little dependence" can be placed upon so few reports from so large a State and only for one season, and moreover most of them were from northern Texas and the Panhandle, where there is always considerable drought. On the other hand, the wiiter knows from personal observa- tion that in the east and south central portions of the State rust has been so detrimental to wheat that this crop has been practically aban- doned over an extensive area of excellent wheat land. It is claimed that at the Texas Agricultural Experiment Station rye is actually injured by leaf rust even in midwinter. In fact, rye is often attacked 21704— No. IC 2 18 CEREAL RUSTS OF THE UNITED STATES. by this rust in all the Southern States. Oats is seldom injured to any extent. Isficaragua wheat, a durum variety, was totally destroyed in 1896 at the Agricultural College farm in Brazos County. The part of the State in which rust seems to be most severe is that lying east of Gainesville, Fort Worth, and Austin. Washmgton and Oregon. — Little information relative to cereal rusts has been obtained from these States, but as far as known they cause only slight damage, except in districts within reach of dampness from the ocean. On account of the growing season being also the dry sea- son, the atmosphere is not usually favorable to rust i^ropagation. Occasionally oats is considerably aftected by the crown rust, as was the case in 189;2 in Sherman County, Oregon. California. — Though reports indicate that rust does not commonly injure the grain, still in some places both wheat and oats are severely damaged. The losses seem to be confined almost wholly to the coast districts within the influence of sea fogs. In other parts of the State' the atmosphere is usually too dry for the rust to spread. In a few dis- tricts along the coast, in parts of Ventura County, for instance, wheat raising is said to have been abandoned on account of rust, but in some districts, as in Mono, Fresno, and San Bernardino counties, rust is extremely rare. Toward the north early wheat is reported by some to be more liable to rust than late, perhaps on account of weakness from winterkilling or frosts. Semihard or soft white wheats are generally raised. Souora is a common variety and is said to be rather rust resistant. Black oats, it is claimed, is more resistant than white. Some of the counties in which rust seems to be especially plentiful are Santa Clara, Ten tura, Sonoma, San Mateo, San Diego, and Santa Barbara. Montana^ Idaho, and Utah. — On account of the cool, dry atmosphere in these States, the rusts seldom occur in sufficient abundance to do any appreciable injury. Virginia. — According to reports, wheat in the Shenandoah Valley is sometimes badly damaged by rust, and oats is damaged quite often. The rust is most abundant in the lowlands. Lancaster wheat is considered as rather rust resistant. Ol'lahoma.—V^^hhm the past three years Oklahoma has become an important wheat producing territory, and in the season of 1897 much injury resulted from rust. There was much rain near harvest time and rust doubtless materially lessened the yield, although it was still very large. In this region May wheat and soft varieties seem to be most subject to rust aud Fulcaster and Turkey most resistant. LOSSES CAUSED BY RUST. Although several writers have discussed the question of financial losses from the grain rusts in difterent countries, including the United States, there are as yet no reliable estimates. Could even an approxi- mate estimate of the losses in this country be made it would be very gratifying, but as yet even this can not be done. It would of course ORANGE LEAF RUST OF WHEAT. 19 be iiiii)ossible to make any general estimate from the per cents of dam- age given in Table 1, as they show the losses for only one or two seasons and in comparatively few localities. In the writer's opinion, however, the average annual loss from rust throughout the Ignited States far exceeds that due to any other enemy, insect or fungous, and often equals those from all others combined. ORANGE LEAF RUST OF WHEAT. {Puccinia rubujo vera tritici). Physiological relations. — It was not until a late date in the writer's inoculation experiments that attention was given to this rust, most of the previous work having been with P. (/ra minis and P. coronata of oats. Many observations made in the field have shown that the leaf rust of wheat and of rye probably do not pass from one host to the other, but are distinct specialized forms. Where wheat and rye are grown in adjoining fields, rust not only attacks them at ditterent dates (which may be partially accounted for in that rye is in a condition most susceptible to rust earlier than wheat, maturing, as it does, earlier), but is occasionally very prevalent on one and not abundant or even present on the other during the entire season. Again, the writer has never been able to find any relation between the abundance of this rust on wheat and its occurrence on native grasses in the vicinity of the wheat fields. The writer's first experiments, reported by Hitchcock and himself (31), pp. 3, 4), showed that uredospores of this rust would not infect oats, rye, or orchard grass, and that a uredo (supposed to belong to P. ruhigo-rera) of Kentucky blue grass would not infect wheat or oats. The results of all subsequent inoculations ^ with this rust are summa- rized in the table following, ' Unless otherwise explained, all inoenlation experiments reported in this bulletin consisted in treating one small pot containing from two to twelve seedling plants of the same kind 10 to 30 days old from the seed in case of cereals and 20 to 40 days in case of other grasses. An average of nine inoculations were made in each experiment, there being never less than six, and rarely more than twelve. The manner of performing the experiment is as follows: The plants to be inoculated are first wet with a very tine spray of water from an atomizer. Then by the use of a thin, narrow-bladed scalpel material is scraped from the diseased plants and is applied in spots here and there on the healthy ones, that is, on the upper surface, the lower surface, or on both, or on the stem also, depending on the particular oxi^eriment. No incisions are ever made. After another spraying the plants are covered with a bell jar, and the latter wet with cold water to aid in preserving a coating of moisture ou the plants. On an average, the bell jar is not removed lor two days, except just for an instant to allow a further spraying with water. During periods of hot sunshine a shade is used to shut off the sun's rays. These experi- ments are made in the greenhouse even during the summer, and the dift'erent rusts are kept growing the year through. All experiments of the same date are, as a rule, accompanied by one check experiment, in which inoculations are made upon plants of the same host as those from which the inoculating material was taken. All inoculations are made with uredospores unless otherwise stated, 20 CEREAL RUSTS OF THE UNITED .STATES. Tablk 2. — Inoculation exjjeriments tvith Uredo ruhujo-rera of nheat. Date of inoculation. Place where experi- ments were made. ' Origin of ! inoculating I material. ' Plant inoculated. Jan. 19, 1897 I Washington, D. C... Do do Do do Feb. 1,1897 do Do do Do do Do do Feb. 12, 1897 I do Do I do Do ' do Feb. 13. 1897. . . . | do Do ■' do Feb. 24, 1897.... I do Wheat. do. do . do . do. do. do. do . do. do . do . do . do . Period of incu- bation (days). Wheat Missogen wheat Rve Wheat Rye Barley Oats.'. Wheat Taganrog wheat Missogen wheat Wheat Dactylis glome rata Akavemidashi wheat Do Feb. 11, 1898'..- Do Do. Do. Do. Do. Do. Do. do Lincoln, Xebr. do do .do do do .do .do -do -do .do -do .do .do -do -do .do California white wheat Elymus canadensis Elyinus canadensis glauci- folius. Elymus virginicus , Agropyron richardsoni .... Agropyron spicattiiii Agrcpjiron tenerum Dactylis glomeruta Panicum autumnale : Result. 8 8 8 8 11 11 11 15 15 11 11 13 7 13 13 13 13 13 13 Successful. Negative. "Do. Successful. Negative. Do. Do. Successful. Do. Do. Do. "Negative. Only one or two spots. Do. Negative. Doubtful. Do. Negative. Do. Do. Do. Do. ' Through inadvertence no check experiment was made. Two of the experiments were doubtful ou account of the slight infec- tions, which were ijossibly accidental. Although under date of Febru- ary 11, 1898, no check experiment of inoculations from wheat to wheat was made, in numerous other exi)erimeiits with wheat alone the inoc- ulations never failed to infect readily, where ordinary cultivated varie- ties of Triticum vulgare were used.^ Although the experiments carried on with Uredo ruhigo-vera were not so numerous as those made with other rusts, they are nevertheless sufficient to show that the limits of this rust form are quite closely cir- cumscribed by physiological conditions. Many more experiments with this rust would have been made during the winter of 1896-97 had it not been for the presence of wheat mildew, which tended to vitiate all work and which could not be got rid of without interfering with the experiments. The time of incubation, or the interval between inoculation and appearance of rust in the case of Credo nibigo-vera of both wheat and rye, is somewhat longer, as a rule, than that of the other cereal rusts, ranging under usual conditions in the greenhouse from seven to ten days. The experiments so far made indicate that only varieties of the fol- lowing species and subspecies^ of the genus Triticum act as hosts for • A fact of interest in this connection is that in one experiment Missogen, a hard ■wheat, of the snbspecies Triticum durum, from Greece, showed no infection. Althongh this may not seem very significant at first, yet if considered in connection with facts shown in Table 8, it "will be found to be in accord with the results of experiments on rust resistance of varieties given in Table 3, which show that durum and poulard ■wheats are more resi-stant to orange leaf rust than are the bread wheats. ■2 The classification of wheats followed in this bulletin is that of Koeruicke and Werner (44). ORANGE LEAF RUST OF WHEAT. 21 tlie orange leaf rust of wheat in this country: Triticnm ridgarc, T. com- j)actum, T. turf/hbini, T. durum, T.spelta, T.dicocciiin, and T. polonknm. These are all cultivated varieties, but varietii^s of the three last named have been grown so far in this country only in an experimental way. Occurrence and dutribution.— It is a well-established fact that the orange leaf rust of wheat is the most common and widely distributed of all the cereal rusts in the United States, and is especially the most constant in occurrence from year to year. The stem rust and the crown rust do not occur on cereals in certain years and certain localities. The orange leaf rust of rye, although occurring wherever rye is growu, is still not common, because rye is not grown generally. Maize rust usually occurs wherever corn is grown, but is seldom abundant; but the orange leaf rust of wheat is not only never absent from the wheat field, being there the year through, but is sometimes abundant even iii dry seasons. There is still so much uncertainty as to the identification of species that as yet it can not be definitely said whether this rust is the most common in foreign countries also, but the probabilities are that it is, at least in all except the countries of northern Europe. In Aus- tralia and India there is no doubt of its being the most common, pro- vided that our leaf rust and the one iu these countries are eciuivalent, which is almost certain. Barclay (5, Vol. XXVIII, p. 257; Vol. XXX, p. -lO) is quite emphatic as to its being the most common in India.' In Sweden the conditions seem to be exceptional. Tlie most common, as well as the most injurious rust in that country, according to Eriksson (31, pp. 208, 209,331), is P. ghmarum, while P. dlspersa is rather insig- nificant both as to its occurrence and tlie injury it causes.^ Wintering of the uredo.—^o much has already been written concerning the wintering of the uredo and the ability of the fungus to readily pass the Aviuter in the uredo stage is so well established for this country that there is little further to add. However, it may perhaps be well to state that the conclusions of Bolley (7, pp. 13-15) and of Hitchcock and the writer (39, pp. 1, 2) have been confirmed and reconfirmed by the writer both in Kansas and Maryland. In the Southern States the leaf rusts of both wheat and rye not only live, but grow all winter. It is now conclusively proved that in latitudes below 40'=^ in this country the leaf rust of wheat is able to pass a perpetual existence in the uredo stage on wheat alone^ without the intervention of any other stage. This is evidently a matter of much economic importance, for if this method is not only possible, but should turn out to be the only one, it is easy to ' Through the kindness of the Government officials of India the writer received specimens of rusted wheat straw from that country. The rust most abundant on the specimens seems to be morphologically enough like our own Fiiccinia riiUfjo-vera to be called the same, specimens showing the second and third stages of this rust being present. However, P. (jramiuis was also found in considerable amount. 2 At the writer's request Dr. Eriksson kindly sent him specimens of both P. ghma- rum and /'. dispersa. An examination of these specimens showed that with the exception of a slight difference in color there is a very close resemblance morpho- logically lietweeen the latter and our F.rubiijo-vera. 22 CEREAL RUSTS OF THE UNITED STATES. see how the fiingus luiglit be easily overcome.' At all events the writer is convinced that the existence of volunteer wheat must have consider- able bearing upon the distribution and propagation of this rust, and the conviction will be strengthened should further experiments fail to reveal other hosts besides wheat. There is still a possibility, however, that some native grass harbors the same rust form. In Australia it is generally accepted as a fact that both Puccinia rubigo-vera and P. graminis exist in the "red rust" stage all the year through, either on self-sown grain or on native grasses. Cobb (19, p. 29) and Lowrie (48, p. 51) make particular references to the matter. However, it should be remembered that identification on morphological grounds alone is not sufficient to prove that the grasses mentioned actually bore these cereal rusts and no mention is made of inoculation experiments. The observations made thus far in India do not indicate that this rust lives the entire year in the uredo stage on any host. Barclay, after careful investigation, failed to discover a continuous repetition of uredospores (5, Vol. XXX, pp. 17, 48), though he admits the possibility of such. He thinks such an occurrence more improbable than "in Europe, where crops both of cereals and wild grasses overlap one another." Prain (00, p. 27) and Watt (70, pp. 51-56) discuss this ques- tion and accept Barclay's conclusions. Eriksson has failed to discover that any one of the wheat rusts lives throughout the year in the uredo stage in Sweden. He claims to have shown, however, that P. glumariim is perennial within the wheat plant (25). To account for this he in^oposes a new theory, which will be ' Dr. Geo. "Watt, editor of tlie Agricultural Ledger of India, makes quite unfavor- able criticisms (70, pp. 55, 56) of the conclusions reached by Hitchcock and the writer concerning this matter, as given on page 9 of onr second report on rusts of grain. After quoting these conclusions he says: "The criticism would seem fairly justifi- able, however, that the.se opinions are based on purely artificial experiments. The experiments may, in fact, be said to show what might occur, not by any means what occurs. It would, of course, be possible (under glass in cold countries or by reduc- ing the temperature in tropical regions) to supply the parasite with continuous crops of its host, and thus to produce uredospores for an indefinite period. * * * The Kansas reports do not seem to afford sufficient evidence in support of the views advanced. « * * In countries that do not have a severe climatic isolation between the sea-sons of wheat cultivation, volunteer survivals of the crop might easily enough occur. Where this is met with the existence of uredospores might be made a matter of actual observation, and their vitality tested at repeated intervals, with- out having to call in the aid of improbable experiments.'" But a more careful read- ing of the bulletin will show that the "opinions"' mentioned are not "based on purely artificial experiments," though greenhouse experiments were indeed employed to confirm field observations. It is distinctly stated on page 2 of the report named that the rust was observed in the field at diff'erent times, up to March 22, and of course after that date rust would continue to grow in that latitude. In the mean- time, to prove that in- these rust spots the spores were actually alive, many of them were brought in at various times and germinated and healthy plauts infected with them, thus "testing their vitality" exactly as Dr. Watt suggests should be done. More than this, diseased plants were occasionally transplanted into the greenhouse, and as fresh uredospores were produced befoie a period of incubation from inocula- tions could have elapsed the mycelia were also shown to be alive (39, p. 2). ORANGE LEAP RUST OF WHEAT. 23 referred to further on. Sorauer (05, p. 21G) states tluit the uredo myce- lium of ruccinia ,str((minis (/*. rnhigo-reni) winters over without injury. The inference is that the locality he has in mind is some jiortiou of Germany, but whether he means P. (jlnmarum or /'. (lispevsa of Eriksson and Heuuings, and whether the host plant is wheat or rye, is not known. LiahHiiy of different varietifs to this rusf. — The study of the compara- tive liability of different wheat varieties to rust has in recent years developed to such an extent that an elaborate paper might be written on this topic alone. To investigate this phase of tlie subject the writer conducted a rather exhaustive series of held experiments during the seasons of 1895, 1S90, and 1S97, and as the orange leaf rust was the one mainly dealt with the question of rust resistance will be discussed at length here. At the outset much time was required to learn what varieties are grown in ditferent parts of the world and the sources through which they might be obtained. Efforts co obtain the varieties of each country were begun early in the spring of 1894, but many of tliose from distant parts did not arrive even in time for jjlanting in the autumn of 1895. Peter Henderson & Co., of New York City, agreed to aid in obtaining foreign varieties through their foreign agents. In addition to those obtained in this way, samples were sent in direct by William Farrer, of New South Wales; Prof, E. ]\I. Shelton, of (»>ueensland ; Prof. S. Tanaka, of the Imperial University, Tokyo, Japan; Carter r>ros., of London, England; and ^Ir. IT. Yagnon, Kherba, Clos des Dras, Algeria. In this country and Canada varieties were obtained from the agricultural experiment stations of jMichigan, North Dakota, Nebraska," Kansas, and New Mexico; from Prof. William Saunders, of the Central Experi- mental Farm, Ottawa, Canada; and from Mr. A. N. Jones, of Newark, N. Y. Yarieties were obtained from every important wheat country in the world. The record of experiments shows over 1,1!00 varieties, but of course many of these were duplicates, and about 75 were Farrer's unfixed crossbreds. Notwithstanding this, however, there were cer- tainly over 900 good varieties tested during the three years of exper- iment. A very large number of these varieties, 4G3 in all, came from New Mexico through Prof. A. E. Blount, who has done so much work in breeding wheats, and who has collected a great many varieties from different countries. These were esi)ecially valuable in that tliey included a number of Blount's crossbreds and also many foreign varieties more or less acclimated to this country. The experiments of 1895 were conducted at Garrett Park, Md., tliose of 189G at Salina, Kans., and those of 1897 at IManhattan, Kans. The soil at Garrett Park was first treated with a fertilizer commonly used for wheat in that region and composed of South Carolina rock, bone dust, and kainit, the application being made at the rate of 500 pounds per acre while the ground was being harrowed. The different varieties were, as a rule, seeded in single alternate rows, these being 48 feet 24 CEREAL RUSTS OF THE UNITED STATES. long" and 12 inches apart at Garrett Park, 20 feet long and 12 inches apart at Salina, and 25 feet long and 20 inches apart at Manhattan. The time of seeding extended from October 17 to November 28, many of the foreign varieties being sown exceedingly late on account of their late arrival in this country. In the experiments at Garrett Park all the varieties, except those from New Mexico and Farrer's crosses, were seeded in two to four rows each. The orange leaf rust was the only species that occurred at Garrett Park, but it was extremely abundant, and therefore there was an excel- lent opportunity of testing the resistance of the different varieties to this rust. After this year it was thought best to continue the experi- ments in the States of the Plains, as they more truly represent the conditions of the greater part of our wheat region, and, moreover, the stem rust as well as the leaf rust is nearly always present in that region. Hence the work was transferred to Kansas. Both rusts attacked the wheat at Salina and at Manhattan, the leaf rust being the more common. At Salina, as at Garrett Park, all varieties were planted rather late because of the drought, the dates of seeding extending from October 16 to November 4. At Manhattan all but about a dozen of the varieties were seeded in good time, that is, on Sei)tember 22 and 23. The seeding of spring varieties was done each year at the average time for such seeding. The winters of 1895-96 and 1896-97 were both unusually severe in the localities where the experiments were conducted, hence hundreds of the imported varieties failed entirely during these two seasons, and only a s^nall proportion of the whole number tested during the three years withstood the climatic rigors and produced seed at the close of the season at Manhattan. It is interesting to note, however, that a number of sorts, especiall^'^ those of Kussian origin, passed the winter at Manhattan in much better condition than most of the varieties reg- ularly grown there by the experiment station (36a, pp. 172-174). The rust liability of the different varieties was determined by noting the per cent of rust on each at the date when the rust was most abundant.' The experiments at Salina were conducted on the farm of Mr. B. B. Stimmel, who kindly gave the necessary ground free of charge. Through the courtesy of the board of regents Qf the State Agricultural College at Manhattan, the farm department cooperated with this Department in carrying on the experiments on its land and furnished certain aid in the prosecution of the work. For the successful accomplishment of the ' If no rust spot could be found on any plant in the row, the per cent of rust was of course zero ; if only one or two spots were i)resent on each of the plants in a half dozen or more places, the amount of rust was estimated as 1, 2, or 3 per cent, and so on ; if nearly all the leaves in the row were about half covered with rust, it was estimated as 50 per cent; and if they were practically covered, it was estimated as 190 per cent. Of course the estimates were necessarily approximate and, moreover, the scale of rusts varies according to individual ideas. In this case the rust was so slight on the sheath that it was not considered. OHAN(iE LEAF KUST OF WHEAT. 25 work at Garrott Pavk great credit is due ^Ir. T. H. T^orsett, of this Division, who rendered valuable assistauce. The principal results of the three years' experiments are arranged in as condensed a manner as possible in the following table. In every case where the rust could not be graded for more than one season the variety is omitted from the table. Nearly every instance of this kind was in the case of a variety that winterkilled or failed otherwise after the first season, but in wsome cases they were simply not at hand for planting more than the one season. As a result of this disposition only about one-third of the varieties actually planted during the three years are represented in the table; hence it shows no varieties from Greece, England, Holland, India, Siberia, Argentina, or Algeria, although they were obtained Irom each of these countries. In the case of the varieties from Argentina and Algeria they were simply received too late to be tested except in the severe season of 1890-07. Only those who have had similar experience can form any idea as to the work required to keep the nomenclature of varieties from being confused. It is not claimed that all the names in the table are right, for such a claim would be preposterous; but it is believed that nearly all are fairly reliable, or at least as correct as they can be in the pres- ent confused condition of our wheat nomenclature. Moreover, many apparent errors are really the correct names, the seeming discrepancy arising from translations of the original name into a different language, or they are cases of pure synonyms; for instance, Kaiser and German Emperor are very probably the same variety, while Kubanka and Beloturka, Black Sea and Arnautka, Thick-set and Cone Kivet, Club wheat and Oregon Club, Genealogic and Hallets Pedigree, Soules, Flint and Yorkshire, German Amber and Amber Winter, De Bordeaux and Eouge Inversible, etc., are all cases of synonyms. The peculiarity of some of the names is doubtless due also to bad orthography, as for example, a certain hybrid is not Tamed but Lamed. In French names incorrect spelling is common and in llussian names it is the rule. Sometimes such simple names as Gold Drop and Blue Stem are applied incorrectly to the most widely different varieties, and such names as Turkey and Russian Red either mean very little or may even be mis- leading, giving no clue to the origin of the varieties as one might sup- pose. A thorough revision of the nomenclature of our varieties of wheat is a most important desideratum for this country. The two main obstacles to a successful comparison of the rustiness of a large number of varieties according to the method followed in the experiments, are (1) that the varieties being necessarily planted at dif- ferent dates those equally susceptible to rust may nevertheless not be equally rusted at the time of grading, and (2) it is impossible to care- fully grade the rustiness of more than one hundred varieties a day. However, in case of the experiments here described, both obstacles were overcome to a considerable degree by grading the varieties each season in the order in which they were sown; but notwithstanding this, many 26 CEREAL RUSTS OF THE UNITED STATES. of tlieui afterwards became more rusted than they were at the date of grading, though this did not occur to any great extent except in 1895. In that year, for example, though a few of the varieties showed no rust at all at the date of grading, yet onlj- one variety, Einkoru, remained absolutely rust proof throughout the season. Much value would have been added to the experiments if they could have been continued in each locality during the three seasons, but under the circumstances this was impracticable. It is hoped, however, that they have settled a few things definitely and that they will furnish a valuable groundwork for future experiments in this country. Tabi.k 3. — Comparative amount of Uredo ruhigo-rera on different varieties of wheat. Per cent of rustiness of w Bald or bearded. varieties in— Aver- age. Localities and naines of varieties of — 1895. 1896. 1897. Remarks. May June June Jnne 23-25. 13-17. 15-16. 22-25. WINTER WHEATS. New Mexico: Algei-iau Xo. \ Bearded .. 2 40 21 Durum wheat. Algerian Ko. 3 American do 3 20 12 Do. do .... x' m* 95 50 55 67 Ames do .... m 50 50 Do. Amethyst Bald X 10 50 30 Blount's cross. Andrews Bearded . . m 92 20 56 Andnis Black do .... 5 15 10 Aowse Bald m 70 50 GO Armstrong do.... X m 88 30 59 Arnolds Hybrid do.... X m 75 25 60 53 * 30 90 20 25 25 58 Baltimore Bald m Banater Bearded . . m 5 20 13 Bennett do.... _ 60 50 55 Bertboud Bald 60 30 45 Big Englisli do :... m 93 10 55 53 Black Cbalf Bearded .. X m 40 35 38 Blue Stem Bald X m 65 40 53 Boyer Bearded .. X 70 50 20 47 Buckeye Bald m 80 '*33 50 54 *Meau of two grades. California Walker Canadian Wonder do • 75 55 60 63 Bearded .. m 92 20 56 Carters H Bald...... X m 5 50 28 Carters I Bearded .. xm 20 60 40 Carters J Bald xm 4 45 25 Champion Amber do.... X 80 35 58 China Red Bearded .. ."iO 25 38 Hard wheat. China Tea do.... 20 55 38 Do. China White do 40 15 60 40 40 50 57 50 37 20 Do. Clawson . . Bald ! Cretan Bearded .. m A durum wheat. Dand Khan No. 2 Bald 5 30 18 White hard wheal. Davis do.... 65 40 53 Deitz Amber Bearded .. 82 60 71 Dera 20 50 35 Early Jasper Bearded . . 60 45 53 Early May Bald lU 60 25 43 97 30 64 Early Kipe ni 88 50 55 64 Earnhardt Bald m 65 50 50 55 Egy])tian F Egyptian No. 1 Emporium do .... 40 55 48 do . . 93 30 62 do .... X 88 35 62 Essex do.... m • 40 *20 30 *Mean of two grades. Fairchild Bearded . - m 98 20 59 Farquhar Bald in 98 50 74 Fleck do .... ni 45 30 38 ' X indicates that such varieties were .just beginning to rust at the date of grading, ^m indicates that the variety was iiadly attacked and sometimes severely injured by mildew {Erysiphe granihiis) at the time of grading. ^Marks of reference (*, t) in this talde call attention to explanations in last coliunn. OKANGE LEAF RUST OF WHEAT. 27 Tablk 3. Comparative amount of Vrcdo rnhifio-rera on lUffirent varieties of wheat- Continued. Localities and names of varieties ol — Per cent of rustiness of varieties in — Bald or bearded. 1895. May June 23-25. 13-17. Bald. WINTER WHEATS— contiiivied. New Mexico — Continued. Kloiirelle I I'luor Spar i tjo Frances do .. Frankenstein *lo Fre»'ling French Black Chaff Fulcaster German Amber German Emperor Gla.ss Gold Drop Golden Chatf Gold Premium Granite Gypsum Gypsy - Hard Austriiuan Hard Manitoba Heiges Prolitic Hicks Prolific No. 1 Hornblende Hungarian Hunters Winter Illinois Second Premium Improved Fife Improved Rice India lied Jacinth x Baart Jennings Johnson Jones Square Head Kings Jubilee x Tourmaline Kivet Lairds Prolitic Landreth Leaks Lehigh ■ Lincoln Little Wonder Longberry Mammoth Martins Amber McCreagan McGees Red Mediterranean Hybrid Mediterranean Spring Mennonite ■ Michigan Amber Midge Proof Millers Prolific Minnesota Fife do.... ....do.... Bearded . . Bald ....do.... ...do.... ....do.... ....do.... do ....do... ....do... Bearded . . Bald do... ....do... do... do... Bearded . Bald do... do ... do... do... Mixed . . . Bald Bearded . Bald do... do... do... do... do... Bearded . Bald do... do... do... .... do... do... Bearded . Bald Bearded . do... Bald do... do... do... Missouri Mediterranean Bearded . Missouri Turkey do — Monarch • do — Moscow do — Nashi do . . . Niagara Bald Nigger Bearded . North Champion Bald Northcotes Amber do ... NoxNo. 3 do... Oakshotts Champion Mixture . Odessa Bald Odessa No. 1 do - . Odessa No. 2 do . . . Odessa No. 3 -, do... Odessa No. 4 1 Bearded . Osterey I Bald Palestine ' do . . . Patagonia do . .. Platinum do . . . ni m m ni m m m m m m m X ra m ra m X m m m m m m m m X m m m m m m m m m m m m m m m m ni m m m ni m m ra m X m m m 1896. 1897 Aver- Remarks. June June; 15-16. 22-25. 95 50 75 60 30 70 i 92 80 , 95 45 60 20 60 30 15 98 60 70 75 75 15 72 60 75 20 85 80 60 98 92 98 50 70 20 95 10 75 95 25 98 15 92 96 60 80 20 20 30 65 98 5 70 65 95 25 10 70 85 35 40 40 55 1 99 97 97 90 95 10 94 40 60 50 50 40 50 55 35 60 50 30 35 55 50 50 50 30 50 25 45 40 55 50 30 35 40 40 45 45 50 50 50 20 50 50 35 30 50 30 40 65 50 *45 30 35 10 50 50 30 60 50 40 25 40 55 50 50 50 15 20 00 10 30 40 55 65 55 50 20 20 60 40 70 50 35 55 45 55 60 65 45 55 65 70 73 50 63 50 40 63 64 60 73 38 55 38 53 40 33 64 55 48 60 58 35 61 45 55 30 63 53 53 74 71 74 35 60 35 65 20 63 63 33 73 33 69 63 48 45 35 58 30 63 74 23 48 53 75 38 40 60 55 28 50 25 43 43 70 76 70 70 58 15 77 55 Blount's cross. Semihard wheat. Blount's cross. Do. Do. Farter's cross. Jones's cross. Farrer's cross. From Australia. * Mean of two grades. Hard wheat. Do. Do. Do. Do. Do. Very hardy. H.ardy. Hard wheat. Hard-grained. Do. Do. A durum wheat. Blount's cross. 28 CEREAL RUSTS OF THE UNITED STATES. Tabi.k 3. — Comparative amount of livedo ruhigo-rera on different varieties of wheat- Contiuued. Bald or bearded. Per cent of rustiness of varieties in— Aver- age. Remarks. Localities and names of varieties of — 1895. 1896. 1897. May! 23-25. June 13-17. June 15-16. June 22-25. WINTER WHEATS — continued. New Mexico— Continued. Pollard Bald 50 40 40 95 5 60 35 55 15 88 94 88 5 40 96 80 80 30 98 5 60 50 92 80 5 65 90 25 94 90 5 93 15 50 96 70 65 85 75 45 05 75 5 91 5 25 40 70 10 75 97 97 50 98 85 75 90 75 50 85 85 96 78 90 96 94 94 90 60 50 60 30 00 30 50 70 "'"so' *20 *4U 50 30 15 30 50 45 35 5 20 50 50 *63 55 50 55 50 45 50 45 50 35 40 50 60 60 05 40 50 50 60 35 45 40 30 20 30 15 60 55 50 55 55 50 10 40 10 30 40 55 15 45 40 20 50 50 45 45 "35' ""55' "'50' "55' "'eo' '"eo" '45' '"50" 75 60 55 45 50 63 33 45 43 63 35 69 57 64 28 35 56 55 65 38 67 3 13 53 50 78 68 28 60 70 35 72 68 28 72 35 45 73 65 63 75 58 48 58 68 20 68 23 28 30 50 8 35 65 74 76 53 74 48 50 58 55 45 62 5(1 71 56 62 69 72 70 68 Blount's cross. Hardy variety. Blount's cross. * Selected bald. * Mean of two gi Hardy sort. Early variety. Flinty-grained. Hard-grained. Hardy and li grained. " Mean of two gi Do. Hardy sort. From Australia. Farmer's cross. Hard wheat. A durum wheat. Hardy wheat. Hardy sort. Early variety. Hardy. Hardy sort. do do.... m m ra X m do T*rincrlpa ^o. 5 .... .......... do.... PrOTJOB ..... ................ do.... do.... Piirnle/ Straw do.... ...""... Bearded .. Mixed do .... m X m m m m Ked May Bald Bearded .. do .... iidcs. Rieti Rietl No. 1 Koberts Mixed Bald do IlOsp worth V Rudy .' Bearded . - do m Rural Bald Bearded .. Bald do ni m m m ii id- Rve "Wheat ades. Scott ...... ........ Bearded .. do.... silver Cliatf Bald do .,. X X m m Small Frame Sdio^ y Bearded . . Bald Tappahaunock do .... m X m m m m m m X m TouzcUe Bearded . . Bald Beardt^d . . Bald Bearded .. do Valley ' VareHotto do Verplauk Bald do .... "Walker Wards Prolific do .... Wards White x Tourmaliue. . W^eek s do .... Bald do .... m White Rose do .... W^hite Rtissiaii Bearded . . do Whites White Velvet Bald do .... m m m m Wicker Bearded . . Bald do .... "W^inter^Teen do .... Witter do .... m ni m X m X m X m X m m m Yellow Missouri do .... Bearded .. Bald do .... 1 do .... Kansas: Badger Big English Buckeye do do .... Dalliis Bearded .. do Deitz Dlehl Egyptian Bald do m X m m X m X m m. 1 m Eai'ly May do Fultz do .... Improved Rice do do .... McPhersou do .... ORANGE LEAF RUST OF WHEAT. 29 Taulk '^.—Com2)araUve amount of Uredo riibigo-vcra on different varieties of wheat- Contiuued. Bald or bearded. Per cent of ruatiness of varieties in— Aver- age. Localities and names of varieties of — 1895. 1896. June 15-16. 1897. June 22-25. Remarks. May 23-25. June 13-17. WINTER WHEATS— continued. Kansas — Continued. Nit^yer Bearded . . Bald Bearded .. Bald Bearded . . do .... X m III III ni 111 III X m X m 111 III m m m m m m m m in X in in X ni m m -x m xm m m in X m 96 5 80 96 88 55 93 94 95 95 75 74 70 70 70 50 92 89 87 95 50 88 91 92 92 88 89 70 91 93 June 21-23 30 30 25 30 70 50 20 45 65 "fio' 64 18 53 63 79 Hardy. Peni Noe d'Umbria ' Bald Durodi Apulia , Bearded . Uuio Riscole do ... Lutaba Mixed ... Majorica Cianco ..| Bald Ma'rzuolo I Bearded Precocissinio di Giappone do ... Prolilero do . . . Rieti I do ... Kieti (prima reproduzione). do ... Taujiarotto do ... Triminia do . . Vareaotto .do . . Vulgo Grauillo e Salerno Mixed . . Misto. France : A Six Raiigs : Bearded Blanc de Flandres Bald Chiildam d'Autonine a Epi do.. Rouge. ' Chiddam de Mars do . . D'Australie Bearded De Noc Bald Hybride de Bordier ' do .. HybrideHatif (i;iiui)au) do .. Petaniellc Xoire de Kice Bearded Ricbelle Blanche de Naples.. . Bald Touzellc Kouge de Provence do . . Victoria d'Automne do . . Spain : l Candeal Desraspado de Mur- ! Bearded cia. Cartagena Rojo Aristado do . . Jejar de Valencia do ■- . Sweden : Kubb ... Russia : Banatka. Bald. Bearded Bearded Winter ao Per cent of rustiness of varieties in— 1895. May 23-25. Champion ■ Crimean ! do Donka Genealogic Red Kalinovka White Bearded Pulavka i do . . Red Bearded ' do . . Red Winter Bald.. Russian-Engli.sl; i do June 28- July 1. (1 30 50 40 50 50 70 50 70 50 GO 70 50 70 80 60 75 60 40 60 10 «0 40 2 65 50 30 40 40 50 65 70 30 60 50 65 70 1896. 1897. June June 15-16.1 22-25. Aver- age. 30 15 30 *50 51 50 50 *30 45 50 45 50 45 60 20 40 50 40 40 40 40 20 50 20 50 30 50 50 40 10 40 15 20 50 50 50 55 20 70 40 40 30 40 50 50 20 50 55 55 60 40 40 *50 50 45 *45 50 *53 60 50 CO *55 60 *65 50 30 33 35 50 51 60 50 53 48 55 58 25 48 65 50 50 63 40 40 50 37 15 65 30 26 48 50 40 20 25 40 33 10 58 60 40 50 45 48 55 35 50 58 48 45 48 53 50 57 58 58 Remarks. Absolutely rust jiroof. Much mildewed in 1895. 'Mean of two grades. Badly mildewed in 1895. Do. Uardy variety. * Mean of two grades. Good, hardy variety. Hardy sort. Poulard wheat. Do. Semihard wheat. Very hard-grained. Hard wheat. Hardy sort and wheat. hard Hardy sort and hard- grained. Hard wheat. Hard wheat. * Mean of two grades. * Mean of two grades. Hard wheat. Hard wheat. "Mean of two grades. Hard-grained. *Mean of two grades. Hard wheat. * Mean of two grades. 32 CEREAL RUSTS OF THE UNITED STATES. Tai5LE 3. — Comparatire amount of Uredo nibi'io-reru on different rarieties of wheat- Continued. Bald or hearded. Per cent of rustiness of varieties in — Aver- age. Localities and names of varieties ol'-^ 1895. 1896. 1807. June 22-25. Remarks. May 23-25. June 28- July June 15-16. WINTER WHEAT — continued. Kussia — Continued. S wedisli 50 *45 *53 *45 60 June 29-30 50 60 50 60 50 50 50 15 60 60 25 20 55 50 15 60 65 60 50 "55 *63 70 65 15 5 10 5 5 50 50 58 58 63 60 60 58 63 58 55 24 10 32 25 4 15 11 55 35 10 33 40 33 Tlieiss Bearded . . "Winter Ghirlsa Bald of three grades. Vyssoko-Lltovslv do ... of two grades. Hardy sort. * Mean of two grades. TTa T(\ - cr r.a i n e d and Tx Bearded . . SPRING WHEATS. liortli Dakota: Glyndon Ko. 669 Bald June 21-23 70 60 65 ■ 65 65 60 6 hardy sort. Gl vndon Ho. 673 do .. Glyndon No. 675 Glyndon Ko. 747 do do Glyndon Xo. 775 do McKissicks Fife do ... Russia: Alsace do Hard wheat ; some Arnautka loose smut in 1895. A durum wheat. Bald 3 5 3 Byelokoloska (Kursk) do 1895. Mildewed in 1895. Chernuska Bearded A durum wheat. Mil- do dewed in 1895. A durum wheat. 1 55 20 5 5 15 5 Badly milde\yed in 1895. Hard wheat ; some Imperial do . Krasnokoloska Bald .... loose smut in 1895. Hard wheat : much Polish mildewed in 1895. Flinty hard. Saxonka Bald Hard wheat; mildewed Spring Ghirka (Ekaterino- slav). Spring Ghirka (Samara) do in 1895. Do. ... do .. Do. 1 A careful study of the table will show what a great variation there is in the ability of the same variety to resist or escape rust under dif- ferent conditions. Some varieties that were almost rust free in 1895 were badly rusted in 1890 and 1897, while others rusted less during the last two years than in 1895. Another important conclusion to be drawn from the experiments is that no ordinary variety of wheat is absolutely rust proof. Even the variety Einkorn above mentioned, which is not really a true wheat, succumbed to the stem rust (P. (jraminis) in 1896. Nevertheless the fact remains that some varieties are better able to escape or resist rust than others under the same conditions and in the same locality, and that some are c^uite resistant, at least comparatively speaking, in all localities and under all conditions, and it behooves the grower to know these varieties. Characteristics of icheats resistant to orange leaf rust. — This is a sub- ject of great interest and has already been discussed to some extent by ORANGE LEAF RUST OF WHEAT. 33 (litiereut writers. So far as the ordinary wheats are concerned, the resistant varieties are as a rule somewhat dwarfed, are close and com- pact, and stool but little. The leaves, comparatively few in number, are stiff, narrow, and erect, with a more or less tough, dry cuticle, often with a glaucous or waxy surface; heads compact and narrow; and grains hard, red, small, and heavy.^ lu other words, the characteristics of these wheats are about the same as those of the wheats of semiarid regions. This is a fact of much significance, being a further argument that more attention should be given to the cultivation of varieties par- ticularly adapted to the black soils and climate of dry steppes, such as those successfully grown in southeastern Eussia, parts of Siberia, and in our own States of the Plains. Fortunately such varieties ])rodu<;e the finest grain and most uutritious Hour known, and are usually hardy, drought-resisting sorts. However, under the influence of certain con- ditions of climate or locality, even these varieties, as above suggested, are likely to prove capricious as regards freedom fiom rust. In fact, no matter what the other conditions, every variety will rust, even con- siderably, if it matures late, and for this reason nearly all the hard Russian varieties mentioned in Table 3 rusted considerably. Early maturity is therefore another important quality. For rust freedom and for other purposes an early-maturing, hard, red, frost-resistant, and drought-resistant winter sort is the ideal one for the greater i^ortion of our wheat region. The foregoing remarks apply only to the varieties of Triticum vulgare and to some extent perhaps to those of T. comjxtctum. But the varie- ties that are almost rust proof, as shown by the table, belong mostly to the subspecies T. durum and T. turgidum^ihowgXi a few of them belong to T. dicocctim, T. monococeum, and T. pohnicum. Those of the first two subspecies, called respectively durums and poulards, although very rich in gluten content, are seldom used in bread making, on account of coarseness and hardness of the grain and the dark color of the flour i^roduced. They are extensively used, however, in the manu- facture of certain pastries and macaroni. Much of the latter is made in this country, and the cultivation of such wheats should be encour- aged. As these wheats are natives of hot, dry countries, they ought to do well, it seems, in the southern part of the Great Plains. If they could be matured sufficiently early (which was not possible in the experiments by this Department) they would certainly be very free from orange leaf rust, which seems to be especially bad in the South. It will be seen that only a few of Farrer's crossbreds are included in the table. The chief reason for this is that those not included simply winterkilled before the conclusion of the three years' experiments, and the amount of seed obtained being exhausted, they could not be tried further. However, nearly all the crosses showed evidence of their ' Varieties resistant to this rust are also often bearded ; therefore in the second •column of Table 3, it is indicated whether varieties are bearded or bald. 21704— No. 16 3 34 CEREAL RUSTS OF THE UNITED STATES. high breeding and careful selection and some proved to be quite rust resistant, but as most of them were unfixed crossbreds their rust resist- ance could be fairly determined only by several years' trial and careful elimination of the most rust liable and less hardy j)lants. As shown by the table, it was found that almost invariably selections from the sporting i)rogeuy of unfixed crosses of bald and bearded varieties pro- duced in the following years plants which showed that the bearded sorts are much more resistant to rust than the bald. In addition to the experiments made by the writer, several were made in cooperation with the Department by i^ersons in different States, the object being to test a few of the same varieties in different localities to determine their rust resistance and adaptability to such regions. Owing to various causes, chiefly lateness of seeding and drought, many of the experiments were entire failures, and in some cases the occurrence of rust Was not carefully recorded. Only four were reported upon, these having been carried, on by Prof. li. 0. Kedzie, at the Michigan Agri- cultural College in 181>G; Mr. B. F. Snyder, at Liberty, Indiana, in 1896; Mr. J. E. Payne, superintendent of the Rainbelt E.xijeriment Station, at Cheyenne Wells, Colorado, in 1897; and by Mr. S. 1. Wilkin, at Bow Creek, Kansas, in 1897. The following table gives a summary of the four reports : Taui,]': i. — Report on Urcdo ruhi The percentages of rustiness are in all cases means of two estimates made by Professor Hays and an assistant. 36 CEREAL RUSTS OF THE UNITED STATES. varieties rusted 8 per cent and under are Haynes Blue Stem,^ Ristings Fife, Glyndou No. 761,^ and Haynes Blue Stem x Glyudon No. 761; tliose rusted 20 per cent and over are Rio Grande, Preston, Percy, Countess, Ladoga, Dawn, Alpha, and Progress. The last seven, which were most rusted of all the varieties, are Professor Sauuders's new crossbred s sent from Canada. Some observations made in the Southern vStates are of special inter- est, because of the usual abundance of orange leaf rust in that region. Tracy (68, pp. 23-25; 69, pp. 44-46) writes that the varieties Defiance and Beloturka, received from Australia, were strongly rust resistant in Mississippi and promise well for that region. Canning Downs, received at the same time, winterkilled. Twelve varieties obtained from England ripened very late and were almost destroyed by rust. Two varieties from France, White Naples and Rieti, made very good yields. In Louisiana, where rust is very abundant, Stubbs (66, pp. 556-561) tested cj.uite a number of varieties at Calhoun and Baton Rouge, most of them having been obtained from California. At Baton Rouge all the varie- ties failed completely. At Calhoun the following varieties were badly damaged by rust: Fonr-rowed Imported Winter. Holhurn Wonder. Red Wheat Gold Finder. Thuring Row. Halletts Red Winter. , Halletts Geuealogic. Hundred Fold. Spanldiugs Prolific. Arizona Indian. Mammoth. Carters Queen. Hybrid Lamed. Common March. Egyptian. Those entirely free from rust or only slightly affected were — Indian Three Months. California Spring. Molds White Winter. Winter Genoese. Russian Red Bearded. Ghirka. Fulcaster. Harris. Big Long Bearded Clul) Brenner. Diehl Mediterranean. Sibleys No. 1. White Russian. Mediterranean. White Boughton. Red Russian. Golden Cross. Purple Straw. Tuscan Island. According to Kellner (41a, pp. 138-143), out of 151 varieties of wheat and spelt tested at the California Agricultural Experiment Station in 1892, only the following were practically rust free: Arizona Indian, Big Long Bearded Club, Blue Grass, F. Gates, INIissogen, Nicaragua, Rus- sian Durum, Russian Red Bearded, Red Sonora, Sicilian, Solid Straw Poulard, White Crimean, and Red Emmer. In Australia the subject of rust resistance has, as already stated, been investigated more, or at least by a greater number of investiga- tors, than in any other country. Several elaborate series of such experiments have been made with difi'ereut varieties and these have ' In duplicate plats rusted as high as 10 per cent. 2 In duplicate plat rusted 11 per cent. ORANGE LEAF RUST OF WHEAT. 37 been reported by I'earsou (57 and 58), McAlpine (51 and 52), Slielton (02 and 03), Lowrie (48 and 49), and others. It is impracticable to quote results from these ditierent writers, but it will perhaps sutlice to give the following list of varieties, which the last International Con- ference on Rust in Wheat' agreed upon as likely to be freest from rust in Australia and at the same time more or less desirable in other respects : (1) Bust-res i.sta7it varieties Wards Prolitic. Marslialls No. 8. Robins Rust-resistant. Marsliiills No. 3. . Anstialian Wonder. For cooler districts Defiance wheats, such as — Wheatons Rust-resistant. Pringles Defiance. Smiths Nonpareil. Blonnts Lanibrigg. Tnnnack. For cooler and moister districts Fife wheats, such as- improved Fife. Hornblende. (2) Varieties eseaping rust on account of early ripening: Allora Spring. Early Para. Early Baart. Budds Early. Canning Downs. Rust-resistant. (3) FroUfic and moderately rust-resistant varieties . Talavera. Leaks. AVhite Lammas. During 1893 and 1894 Maddox (50, pp. 14-23) made observations at Eastfield, Tasmania, on the occurrence of rust on over two hundred varieties, including seventy-five or more crossbreds from Farrer. The following varieties and also 28 of Farrer's crossbreds were free or pra(jtically free from rust : Wards Prolific (white). Manitoba. Wards Prolific (red). Pringles No. 5. Bearded Herisson. D'Arblays Hungarian. Bald Herisson. Improved Fife. Fultz. Niagara. Summer Club. Robins Rust-resi.stant. Hornblende. Medeah. The following varieties were free, or practically free from rust during 1893 and 1894: Anglo-Australian. Blounts Fife. White Fife. Anglo-Canadian. Bega. Tourmaline. Sicilian Square-headed Red. Saskatchewan. I Agr. Gaz., July, 1896, Vol. VII, pp. 4.38-442. 38 CEREAL RUSTS OF THE UNITED STATES. Two of Farrer's crosses and the following other varieties were rusted 50 per cent in 1894 : Bellevne Talavera. Kings Jubilee Quartz. Clawson. Hudsons Early Purple Straw. Lazistan. Rattling Jack. Frames Early. American Champion Head. Red Tuscan. Victorian Defiance. High Grade. White Naples. Urtoba. Early Baart. Cape Wheat. California Spring. The following- varieties were rusted 75 per cent in 1893 : Square-head. Red St. Laud. Count Walderdorffs. Frames Early. Carters Stand-up. Red Tuscan. Boutchers Velvet. High Grade. Golden Drop. Urtoba. Bestehorns. Cape Wheat. In Sweden, according to Eriksson (31, pp. 340,341), there was until recently but little difference noted in varieties as regards liability to brown rust (Puccinia dispersa), although his experiments showed great differences with respect to yellow rust (P. glumarum). However, in 1890, when the brown rust was unusually abundant and severe in that country, and when the teleutospores were for the first time observed on the stems (sheaths), at least to any great extent, there was a great dif- ference in the amount of this rust on the different varieties. In another account of the rust that year (27, pp. 248-251) Eriksson divides the varieties tested into three classes, according to their degree of rustijiess. Class 1 includes the varieties free from rust, at least as far as the sheaths are concerned; class 2 includes the varieties sparingly or con- siderably rusted ; and class 3 includes those badly rusted. Examples of class 1 are — Horsefords Pearl. Michigan Bronze. Scholeys Square-head. Kinver Square head. Ble a Epi Carre. Count Walderdorlfs. Examples of class 3 are- Frankenstein. Manchester. Grevenhagen. Shireft's Square-head. Hickling. Blood Red. Red Chafi" Danzig. Beselers Brown Thick-head. Hungarian White. White Club Spelt. In India very little has been demonstrated concerning the liability of different varieties to rust, though various experiments with wheat have been conducted for many years in different provinces of that country. In the Central Province, where P. rubigo-vera, or " gerwa," as the natives ORANGE LEAF RUST OF WHEAT. 39 call it, is said to be the i)riiic'ipiil disease with which farmers have to couteiid, the wheats geueially regarded as rust i)roof are Muiidia IMssi and Baiisi, but the former is not a very marketable sort. xVt the Nagpur Experimental Farm experiments were made in 1895 (12, pp. 19-21) to test the comparative liability to rust of four varieties, includ- ing Mundia Pissi, and a similar series was planned in 1897 (13, pi). 22-24) with eight varieties, including ^lundia I'issi and Bansi. As but little rust occurred on any of these varieties experimented with, no conclusions could be drawn. Twenty-nine of Farrer's crosses were tried in 1897. The 2G of these that came to maturity were very late in ripening and rusted badly, though no other varieties were rusted. The rust, liowever, did not injure the grain. A variety, supposed to be from America, was tested at theCawnpore Experimental Farm in 1890 (.')7, pp. 22, 23) and remained entirely free from rust, while the conunon sort, Muzaffarnagar, beside which it was grown, rusted here and on other lields of the farm. Judging from the description of the variety, however, it is probably not of American origin, but a durum wheat, perhaps from the Mediterranean region. Bust-free varieties for the United States. — Before leaving this subject some suggestions may be given concerning rust-free varieties for this country. By reference to Table 3 it will be seen that after discarding all that are not true bread-wheat varieties there is really little differ- ence, so far as rust is concerned, in the sorts tested by this Department in 189G and 1897 or in many of those tested in 189.5. This is due chietly, if not wholly, to the late seeding. ^Nevertheless, in case of most of the varieties in 1895, and even in the next two years, if the maximum of rustiness to be tolerated is raised considerably (to about 40 per cent), a number of bread wheats remain fairly rust resistant. Judging from all the experiments Jind observations discussed, the fol- lowing varieties, already well known and good standard sorts in other directions, ma}^ be recommended as likely to prove considerably resist- ant to orange leaf rust in every part of this country, provided, of course, that they are sown in time : Winter wheats: Turkey. Rieti. Mennonite. Odessa. Pringles No. 5. Pringlea Defiance. Spring wheats": Hayiies Blue Stem. Saskatchewan Fife. The following varieties seem to be resistant, but have not yet been well established as such : Theiss. Fulcaster. Oregon Club. Deitz Longberry. Sonora. • Arnolds Hybrid. Diehl Mediterranean. California Spring.' 'Although according to name a spring variety, it withstands the winter quite well. 40 CEREAL RUSTS OF THE UNITED STATES. Some of the liardy prolific sorts not yet well known in this country, but likely to be more or less rust resistant after thorough acclimation and selection, are — Winter wheats : Prolifero. Winter Gliirka. Banatka. Budapest. Red Winter. Crimean. Nashi. Yx. Tangarotto. Belleviie Talavera. Bearded Winter. Spring Avheats: Alsace. Spring Ghirka. Two varieties which are quite susceptible to rust, but which usually ripen early enough to escape the worst effects of it, are — Early May. Zimmerman. Some others not quite so well known, but probably worthy of trial as rust-escaping sorts, are — Early Baart. Japanese No. 2. Allora Spring. Yemide. Kathia. Canning Downs. Rosewortliy. These last varieties, however, are not likely to withstand very severe winters, and are therefore best adapted to southern districts, where they may perhaps in time become acclimated. Yemide and Kathia are probably the most hardy of the six. As already stated, Canning Downs winterkilled in one trial even in Mississippi. The following durum and poulard wheats adapted to hot districts and under most conditions extremely resistant to the orange leaf rust, are suggested as being well worthy of trial as macaroni wheats in the Southern States. Aruautka. Petanielle Noire de Nice. Taganrog. Gallauds Hybrid. Beloturka. Cherniiska. Nicaragua. Cretan. Medeali. Missogen. Two varieties of the emmer and monococcuni groups from Germany that are recommended for further trial as stock leed are — Einkorn. Black Velvet Couipouiid Eiiimer. These are very resistant to rust, Einkorn remaining absolutely proof against the orange leaf rust during two years when it was very severe. Damage.— In previous publications (16 and 17, p. 497) the writer expressed his opinion that after all the orange leaf rust, as a rule, does very little, if any, damage to wheat, even during periods when it is quite abundant. Although further investigations have confirmed this belief, he is now inclined to think that occasionally, under certain OKANGE LEAF RUST OF WHEAT. 41 conditions and in certain localities, considorablo injury may ensue if the rust occurs mucli in advance of harvest. In all instances, however, where the wheat is a total loss, which often occurs in such States as Kentucky, Indiana, Texas, Michigan, and Ohio, the black stem rust is found to be the chief, if not the only, rust present; and in all the writer's experience he has not met with a single well-authenticated case in this country where the leaf rust caused actual shriveling of the grain.' In recent years botanists of all countries except England and Sweden seem to agree that F. ruhi(jo-vera is not only the common rust, which is true, but also the one of economic importance, as will be seen by con- sulting the writings of Arthur (1), Bolley (7, pp. 13, 14), Barclay (5, Vol. XXX, pp. 4G, 47), and Oobb (1 13 9 40 10 20 20 10 3 10 5 4U a2 10 10 12 40 13 12 8 40 12 10 8 12 5 Saiina, Juneyo, 1896. 90 85 90 90 00 85 95 80 90 91 94 94 94 90 91 90 80 95 85 92 92 94 95 I 91 90 ! 90 ' 95 : 94 , 92 ! 93 I 85 ' 85 . 95 I 90 ; 70 i 90 i 92 Avorajfe. 90 88 92 85 90 90 96 93 94 96 4 85 45 5 *93 49 3 92 48 4 92 48 4 90 47 3 90 47 5 92 49 4 90 47 5 94 50 3 92 48 1 85 43 4 93 49 3 90 47 5 94 50 4 94 49 5 *05 50 4 92 48 3 91 47 4 *90 47 1 80 41 1 80 41 4 91 48 a 92 49 Kcniarky "'0 47 50 50 50 46 54 44 50 51 53 53 I 54! 50 : 06 ' 50 50 58 ' 48 i 49 51 50 68 ; 52 I 50 50 I 54 I 67 ' 53 I 53 I 45 ' 63 ! 54 50 39 51 49 47 45 48 44 47 47 51 49 50 51 .Small grain and .«hort straw. Same as Georgian oat. Very tall and heavy straw. •Mean of two grades. Late, but drought resistant. Rather large grains. *Mean of two grades. Do. Same as Georgian. A very early variety. 52 CEREAL RUSTS OF THE UNITED STATES. Tablk 7. — I'redo coronata on rarietles of oats — Ci>iitiinied. Localities autl names of varieties. |Subspe- cies. Per cint of rust i- ness at — Garrett ' Salina, Park, June 30, July 16. 1895, 1896 Average. Kemarks. France : Belgian Black Winter Black Hungarian s DEtanipes p Grey Houdan p Grey "Winter p Jaune de Flandre p Joanet*e p Noire de Brie p Rousse Couronnee p "White Hungarian s Siberia : Siberian (Nerchinsk) Siberian No. 7 Spain : Civada Blanca Grjuija de p Barcelona. Portugal : Abrantes p 93 48 95 49 j 94 48 1 95 49 88 45 90 47 92 48 88 45 92 48 92 48 70 70 60 60 95 95 88 45 1 A black chart" sort. The Flemish oat. Short straw. * Not planted in 1895. Do. Tliere was little difference in the amount of rust on different varieties during the second season, all being very badly affected. As the black stem rust also was present, it was very diflScult to grade, even approxi- mately, the amount of crown rust alone. It is believed that had the varieties been sown quite early in the case of the experiment at Gar- rett Park many of them would have escaped entirely, as the rust did not appear until very near harvest time, and only four varieties were rusted as much as 40 per cent. On the whole, the two years' experi- ments do not show much difference in the comparative liability of varieties of oats to the crown rust. Damage. — The writer knows of no instance where it could be proved that this rust cau.sed any really serious damage to oats, under ordinary conditions, when the black stem rust was not also present; but never- theless it is quite possible that it may occasionally cause considerable injury. In any event it seems to be of more economic importance in this than in most other countries. BLACK STEM RUST OF WHEAT. [Puccinia grammis tritivi Eriks. and Henu.). Pnccinia granipiis has always served as a convenient species in which to include such grass rusts as show no particular morphological indi- viduality, but which resemble this species, and as a result it is credited with more host plants than any other rust. The writer has found at least 38 different grasses given as hosts for this rust in America, and Eriksson states (31, p. 119) that lO.j are recorded in the herbarium of the division of plant physiology of the experiment station at Stock- holm as hosts for this species in Sweden. Xearly all such identifica- tions are of little value from an economic standpoint, many of them are wrong morphologically, and some of them are little better than guesswork. Errors in identification often arise from the fact that the specimens show but one stage of the rust, usually the uredo stage. BLACK STKM RTTRT OF WTIKAT. 53 The miinber of host phmts has beeu considerably reduced in recent years by the separation of a number of new species, and also by more correct identifications. Some of the new species occurring in this country are P. aniphu/cna Diet., T. subuiteiis Diet., P. <1i,sfichlitJis Ell. and Ever., P. at/ropyri 1011. and Ever., and P. jiihnfa Ell. and T>arth., but whether these are all good species the writer is not yet prepared to say. The last one, however, he believes is not.' PhyKiohH/ical relations. — Long before the writer's inoculation experi- ments were begun he often noticed, as doubtless others have also, a considerable difference between the general appearance of the black stem rust of wheat and of oats. In fact tlie uredo stages on the two hosts, at least on the leaves, ai)pear to differ from each other more than the uredo stages of /'. rubigo-vera and /'. coronata. The sori are longer, larger, and seemingly a little darker on oats than on wheat, and usually there is proportionally more of the uredo present on the former than on the latter, though this latter feature is quite variable. Of course such differences maybe due to differences in constitution of the host plants, but this is not necessarily so, for the leaf rust and crown rust on the same hosts behave differently. Believing such work would be of particular interest, the writer inaugurated some inoculation experiments with this widely distributed rust species, the experiments with the form on oats being started sooner and carried on to a greater extent than those with the form on wheat, which latter, on account of lack of inoculating material, were not begun until Aj^ril 20, 1896. The sources of inoculating material were Uredo (jramlniH of wheat, kindly sent in by Prof. E. II. Price, April 14, 1896, from the Texas Agricultural College farm, and barberry rust, furnished through the kindness of I*rof. C. F. Wheeler, May 9 and 11, 1890, from the Michigan Agricultural College. These experiments are very incom- plete, but are still under way. The following table gives a summary of the work so far done : Tablk 8. — Inoculation experiments with Uredo Do... Do... Do... Do... Do... Do... Oct. U Do... Do... Oct. 21 Do.. ...do ... ...do ... ...do ... ...do... ...do ... ...do ...! ...do ... ...do ...' ...do ...; Agropy- j run (e- nenuii. \ do . . . I Wheat Do do Nov. C do do do .. -do do ... .do I I'.arley .. do lliirdeuin juba- tU'll. ...do ... ...do... ...do ... ...do ... ...do ... ...do ... ...do ... ...do ... . . . .do . . . Barley - ■ ....do ... .do . .do . -do . .do , .do . .do , .do. .do .do .do .do Xov. 12 Do... Do. Do. Do. 1898. Jan. 5 Do... Do-- Do.. Do.. Jan. 21 Do.. Feb. 11 .do . .do. .do . -do. .do. Lincoln, Nebr .do. .do. .do. .do, do -do .do Do... Do... Do... Do... Do... Do... Do... Do... Do.. Do.. Do.. Feb. 21 Feb. 25 Feb. 28 .do. .do. .do. .do. -do. -do. -do. -do. .do. .do. .do. .do. .do. .do. ....do . llordeum jub a- turn. ....do ... ..do... .do, .do .do Wheat Rye Barley Wheat do A ka veinidii.sbi wheat do Einkoru (Triticum monococeum) Ely lit us virginieu.t Wheat .do. .do. Barley Oats... Barley Rye... Wheat Harley Oats Arrhenatherum claims Triticiim villosum Dacti/lis i/loinerata Wheat ..' Barley Coin Wheat Barley Hordeum nodosum Wheat Wheat Barley Rye Oats Hordeum nodotum . Elymus ca nu- de 71 sis ylauci- folius. ...do... ...do... ...do... ...do... Wheat.. ...do... Elymus cana- d e nsis (jlauci- /alius. ....do... ....do... ....do... ....do... ....do... ....do ... ....do... ....do... ....do... ....do... ....do... ....do... ....do... ...do... Wheat , Barley Oats Rye Agropyron richardsoni . Wheat Barley Wheat Barley Rye :-• Oats — Elymus canadeyisis glaucifolius Elymus virginicus Agropyron teneruin, Agropyron spicatum Elymus hirsutiglumis Elymus virginicus muticus DactyUs gloinerata Agrcstis alba vulgaris Elymus canadensis Ilordfum j ubaturn Wheat 12 12 lU 16 Ifi 16 16 16 16 13 13 12 13 8 8 8 10 U 11 11 11 11 11 11 11 11 11 16 10 16 6 11 11 11 11 11 21 21 21 21 21 10 10 14 Sncces.sfnl. Only one or two spots. Negative. Do. Do. Do. Do. Successful. Do. Do. Negative. t)nly one or two spots. Negative. Only one or two .spots. Sucie.ssful. Doubtful. Successful. Negative. Successful. Do. Negative. Do. Do. Do. Successful. Do. Negative. Doubtful. Only ouo or two spots. Negative. Successful. Successful. Oulj' one or two spots. Negative. Do. Do. Successful. Do. Negative. ■Do. Do. Successful. Do. Do. u Do. 14 Negative. 14 Do. 14 Successful 14 Negative. 14 Do. 14 Do. U Do. 14 Do. 14 Do. 14 Do. 7 Successful 11 Do. 9 Do. 56 CEREAL RUSTS OF THE UNITED STATES. The results above given show that as a rule this rust form passes readily from wheat to barley and from barley to wheat, the exceptions not being more numerous than the faihires from wheat to wheat, and such exceptions were probably generally caused by the presence of mil- dew or some other unusual condition. Indeed, it usually seemed that the transfers from wheat to barley and from barley to wheat, especially the latter, resulted more easily than did transfers from wheat to wheat or from barley to barley. As shown by the writer's investigations, it is the only cereal rust that readily infects more than one of the five cereals. In a few instances there were apparently slight infections of oats, but it was evident, if they were not accidental from Uredo graminis of oats, that the infections took place with difficulty. One interesting feature of the experiments is that although the bar- berry rust in one set of experiments readily infected barley, it failed to infect wheat or oats and produced but one spot on rye. If the bar- berry rust was originally produced by the sporidia of teleutospores from wheat, which is not known, the failure to infect oats would be in accord with Eriksson^s experiments (20, pp. 304, 305; 29, pp. 200-202), which showed that rust of barberry would infect only that cereal the rust form of which originally produced the barberry rust. On the other hand, the fact that it also failed to infect wheat, and infected rye but slightly, . can not be thus exj)lained; and this, with a few other peculiar results obtained, has led the writer to suspect that in this country there may be two distinct forms of the black stem rust on barley, one of which also infects rye but not wheat, and the other wheat but not rye. Like barley, Hordenm juhatum seems to also act as a host for both forms. Similarly peculiar results were obtained with certain other wild grasses and need to be further invCvStigated. As previously stated, the leaf rust seemed to infect the true bread wheats more readily than it did the durums and poulards, but in these experiments the reverse was true, the stem rust infecting durums and poulards more easily than it did the bread wheats. Moreover, larger and darker sori seemed to be formed on the durums and x)oulards than on the other wheats (PI. IV, fig. IS) — an interesting fact in connection with the subject of rust resistance of varieties. For the inoculations from the wild grasses Hordtum juhatum, Agro- pyron teneruin^ and Elymus canadensis glaucifoliuSj material was of course obtained in the field on these hosts. Two rusts are found on A. ienernm^ both of which seem to produce the same sort of sori on wheat, but only one is believed to be P. gyam'mis. The spore forms were quite different. Under ordinarj^ greenhouse conditions the i)eriod of incubation for this rust in the uredo ranges from eight to twelve days. Under dates of September 2G, 1896, November 6, 1897, and February 21, 1898, the shorter periods noted include only the time until signs of rust appeared, the epidermis not being broken until one or two days afterwards. From experiments and observations so far made, the following may BLACK STEM RUST Ol^^ WHEAT. 57 be considered as established hosts for the black stem rust of wheat in thiscountry: THticumvaUfare, T. compact urn, T.turgichim, T.actylis ' , ijlo m ■ I I erata. . do do . .. 7 '[ Manhattan, Kaus .do . -. Plant inoculated. Fen tons Rust Proof oats. Loliinn jterenne A rrhenatherum elatius . . . Hordeitin jubatuni Bromus tmioloides Avena pratensis Agrostis alba vulgaris. Phleum pratense Lolium perenne Hordeitm mttrinum Dactylis glomerata Shirosawa wheat Trisetum subspicahiiti Alopecurus al2>estris Holcvs mollis Agrostis scabra Fentons Rust Proof oats- Polyj)ogon monspeliensis. . Feetuca sp. indet JEatonia sp. indet JBrizopyron siculum Sporobolus asper Poa pratensis Androptigon halapiertse . . . Phalaris arunduiacea Anthoxanthvin odoratuni . Ligo wo oats Triticttm spelta centivum.. Sporobolus cryptandrus.. . Panicum crusgalli Hordeum murinum Australian Glory wheat Kotleria cristata A inniophila arenaria Dun oats Avena. iterilis Phleiim aspervm A m in oph ila arenaria Eatonia dudleyi Panicum crus-galli Poa annua Brachy podium distachys Sporobolus cryptandrus Eleusine egyptiaca Bouteloun curtipendula Eragrostis pursldi Schedonnardus 2)anienlattis . Triodia cuprea Oats do Barley Rye Wheat Oats do Aveiia 'pratensis ........... Avena hookeri Dactylis glomerata Eatonia obtusata Bronui.s ciliattis Tritic u m villcsu ni Hurdeu m nodosum Avena fatua Festuca ovina Festuca arundinacea Oats Festuca rubra glaucescens . Festuca gigantea Wheat Oats . . . Wheat . BLACK STEM RUST OF OATS. 63 Tablk 10. — Inoculation experimeHta with Uredo (/raminis avencv — Continued. Date ofinocii- latiun. riaco where experi- lueuts were made. 1897. Sept. 7 Do... Dec. n Orijiin j of inocu- lating material. Manhattau, Kaiis . do I^incoln, Nebr Do. .do VaclyUs glom- erata. ...do ... Arrhen- athc ■ r u tn el al- ius. ...do ... Plant inoculated. Period of incu- bation (day.s). Oats Dactylis glomerata . 0at3 Rye. 1898. Jan. 5 Do... Jan. 24 Do. Do. .do .do .do .do .do Oafs . ...do ...do Oats ... Wheat Data ... .do .do Wheat Oats .. 8 12 12 21 21 12 12 9 Result. Successful. Do. Do. Negative. Successful. Ncnjative. Ouly one or two spots. Negative. Successful. Inoculations with this rust were successful on a greater number of hosts than inoculations with any other rust. Of course, as in the case of Uredo coronata, some of the e.xperimeuts might not have proved successful with older plants, but nevertheless a number of inoculations on older jdants also produced infections. On the other hand, in the case of Etampes oats, under date of March 12, 18!)0, and in a few other cases, the plants were inoculatedjust after germination upon damp cot- ton, but evidently were too young to be affected, the germ threads of the spores having ceased to grow before the first leaf was accessible. An interesting feature of these experiments is the number of in- fected plants that belong to species of allied genera or to species that are rather similar in structure, and their close correspondence with those infected with Uredo coronata, as shown in Table 6. The first forage grass proved to be a host for this rust in this coun- try was Dactylis glomeratct. On account of its susceptibility to infec- tion, both when young and in an advanced stage, the writer for some time believed it to be a host for this rust. On August 2, 1897, both uredospores and teleutospores of the rust were found in great abund- ance on this grass on the grounds of the agricultural college at Still- water, Okla.' A rusted i^lant was transferred from these grounds to the greenhouse of tlie Agricultural College at Manhattan, Kans. Material from this plant easily infected other plants of orchard grass and oats, producing sori on oats more vigorous than those from mate- rial of the same rust taken from oats. The rust was found on Arrhena- theruni elatlus at Lincolu November 16, 1897, and the reverse infectious on oats from this material were produced December 23, 1897. On the basis of experiments and observations made so far the follow- ing may be named as well-established hosts for the black stem rust of ' He had, of course, seen the rust on orchard gra.ss several times before, as it is rather common, but had always supposed it to be e(|uivalent to the form on wheat. 64 CEREAL RUSTS OF THE UNITED STATES. oats : Avena sativa patula, A. sativa orientaUs, aud A.sativa nuila — culti- vated varieties, aud DactyJis (jlomerata, aud Arrhenatherum elatius. The followiug species may at preseut be cousidered as probable hosts : Avena fatna, A. hoolceri^ A. pratensis, A. sterilis, Koeleria cristata, aud Lolium perenne. Occurrence and distribution. — The black stem rust of oats is certainly- more common thau the black stem rust of wheat, owiug probably to the fact that oats matures later than wheat and the stem rusts occur later than the other rusts. However, it is not known whether the stem rust of oats is generally commoner thau the crown rust, although) definite statements on this point might be made in the case of certain localities — locality appareutly being an important feature in this connection. As in the case of stem rust of wheat, this rust is not constant in occurrence, although perhaps more so thau the former. Reports indi- cate that in the Southern States it is neither so common as it is in the Northern States nor so common as the crown rust. iSTo trace of it was seen in Maryland during two successive seasons of observations by the writer. His examination of old straw at the Texas Agricultural Col- lege in December, 1895, showed that the crown rust was quite abundant the j)receding summer, but none of the stem rust was observed. Wherever oats is grown iu foreign countries the distribution of the stem rust of oats seems to be quite general, but in some regions it is said to be of little economic importance. Wintering of the uredo. — So far as yet known, this rust does not win- ter its uredo in this country. The following table gives a summary of the writer's observations of the occurrence of the uredo iu late summer and autumn : Taulio 11. — Snmmanj of observatiouti on late ucciirrenre of Uredo f/raminis arcnce. Host plaut. Date of observation. Place of observation. Oats Sept. 17, 1896 Nov. 2,1896 Aug. 27,1897 Aug. 23, 1897 Oct. 8, 1897 Oct. 12,1897 Nov. 12, 1897 Nov. 14,1897 Nov. 16,1897 Manhattan. Kan.s. Do. Do. Stillwater, Okla. Do. Payne County, Okla. ; Manliattan. Kaus. Lincoln, Nebr. Do. Do Do Dacti/iis qlomerata.. Do . .' Oats Daciylis (jlomerata.. Oats Arrhenatherum ela- tius. The fact that the uredo was found on other hosts than oats, and espe- cially the fact that .these hosts are perennial, strengthens the probability that the uredo winters over, but at the same time weakens the proba- bility that it winters on oats. At any rate, it is easily seen that orchard grass and tall meadow oat grass are not likely to prove desirable neighbors for the oat field. As already stated, Cobb (19, p. 29) mentions that P. graminis winters over in the uredo stage iu Australia, but he is not sufficiently explicit to justify the couclusion that he includes this form on oats. BLACK STEM liVST OF OATS. 65 Liability of different varieties to this rust. — But litth>- iulbrmution exists as to the liability of different varieties of oats to rust iu this eountry >>ased on any case in wliioli it was detiuitely known that the stem rust was the chief one under observation. In the season of 1894 the writer assisted Prof. J. IJ. Shepperd in grading" the rustiuess of varieties of oats grown at the North Dakota Agricultural Experiment Station, F. graminis being tlie dominant rust on oats that season, not only there but also throughout the North and West. The following- season Prof. H. L. Bolley, of that station, graded the rustiness of the same varieties, and in a recent letter received from him he states that P. (jramlniH was the chief rust present that year, but that the presence of P. coronata accounts for the small percentages of rust noted for cer- tain varieties. The results of the two years' grading were reported by Professor Shepperd iu a bulletin issued March, 189G (04, pp. 38-40). The i^erceutages for the same varieties for the two seasons correspond quite well as a rule, antl show almost entire freedom from rust in the case of Early White Kussian, Fentons Rust Proof, Tartarian, and Great Northern, and an abundance of rust on Giant Yellow, American White, Badger (^ueeu, and North Star. This rust occurred to a small extent on the varieties of oats in the field experiments at Salina, but on account of its late appearance and the great abundance of the crown rust it was impracticable to grade the amounts on the different varieties. In Australia, where so much has been written on rust-resistaut wheats, apparently no attention has been given to rust-resistant oats, probably for the reason that oat rust is of little economic importance in that country. In 1803 Eriksson made some observations (31, pp. 348- 350) on rust liability of varieties of oats in Sweden, es])ecially as regards the black stem rust, and concluded that iu Sweden there is no particu- lar difference in the susceptibility of different sorts of oats to this rust-. Damage.— The damage which this rust causes to oats iu this country is even greater than that produced by the form occurring on wheat. In fact almost every year thousands of acres are totally destroyed in some portions of the country by this parasite, the damage being great- est north of the thirty-seventh degree of north latitude and east of the ninety-fifth degree of west longitude. According to Eriksson (31, p. 386), it is also very destructive at times in Sweden, the damage to oats in that country from this cause in 1889 amounting to IG million crowns (about 4.^ million dollars). MAIZE BUST. {Puccinia, sorghi Schw.). Physiological relations, — The fact that the maize rust bears some mor- phological resemblance to the rusts of certain grasses, particularly to P. graminis of wheat and oats, led the writer to suspect that it might ' 21704— No. IG 5 u CEREAL RUSTS OF THE UjstiteD STATES. be equivalent to one or more of these, and it Avas for tliis reason espe- cially that in the inoculation experiments with the rusts of the other four cereals Indian corn Avas employed several times, as were also other grasses. A few experiments were also made with the uredo of maize rust itself. The following results were obtained: Table 12. — Inocidalion experiments with Uredo sorghi. Date of inocu- lation. 1896. Sept. 17 Sept. 21 Dec. 22 1897. • Apr. 10 Oct. 13 Place where experi- ments were made. Origin of inoculating material. Plants inoculated. Manhattan, Kana j Corn Corn . . do 1... -do do. Washington, D. C- ' do do . do I do Euchlceiia mexicana. Manhattan, Kans ] Euchl cena \ Corn inexicaiia. i Period of incu- bation liesult. (days). 5 Successful. 7 Do. 1 Do. 11 Do. U Do. The time of incubation for this uredo is shorter than that of any other cereal rust, varying from five to eight days, under ordinary greenhouse conditions. Until recently it was not known that maize rustioccurred on any host in this country except Indian corn. In pursuance of an idea long entertained that teosinte {Euclilccna mexicana)^ on account of its close relation to corn, might also be a host for the maize rust, the writer planted some teosinte seed (obtained from the Division of Agrostology, which kindly supplied seeds of various grasses used in the writer's experiments) in tlie greenhouse in the spring of 1897, and successfully inoculated the resulting jjlants, a longer- period of incubation being required, however, than in the case of corn. On August 21, 1897, he found both the uredo and teleuto stages on teosinte growing on the grounds of the Agricultural College at Stillwater, Okla., and later suc- cessfully inoculated corn with uredospores from the teosinte, vigorous sori being produced. It is a curious fact that while inoculations with this rust resulted in adding another host for it, similar experiments with the other cereal rusts have shown that the number of their hosts is much smaller than had been. supposed. The established hosts for this rust in the United States, as determined by experiments so far made, are Zea mays (cultivated varieties of the groups saccharata, dentiformis, and vulgaris) and Euchla'ua mexicana. Occurrence and distribution. — Maize rust occurs in all parts of the United States where the ordinary varieties of field corn, sugar corn, and pop corn are grown, but it is rarely found in great abundance in any locality. It attains its maximum abundance on ordinary field crops about October 1, or about three months later than any other rust of cereals. Whether the introduction of teosinte as a forage crop will have any effect iii facilitating the spread of the rust remains to be seen. MAIZE RUST. 67 This rust is also reported from South America, Germany, France, Italy, Portugal, India, and south Africa; but owing to the facts that the genus Sorghum is included with its hosts and that Barclay has shown (5, Vol. XXVIII, pp. 257-259; 3, pp. 21-1, 215) that the rust ou Soryhiim vulgare in India is a distinct species, whicli he called P. pennisetij it is ques- tionable whether it actually occurs in all these countries. Wintering of the iiredo. — This rust is not yet known to winter its uredo in any country, and as no iccidial stage has been discovered, the manner of its i^erpetuation throughout the year is entirely unknown. Liahility of iJ iff event varieties to this rust. — There is as yet no differ- ence known in the susceptibility of different varieties of corn to this rust, but so far no investigations have been made to determine this point. The rust seems to be of but little economic import-ance. Damage. — The writer has occasionally seen patches of very late corn injured by this rust. If corn is sown broadcast as a forage crop late in the season it is likely to bo badly damaged, but as a rule the rust does little or no damage. Pammel mentions (oOa, p. 854) that in Iowa the yield of corn, especially of sweet corn, is sometimes materially lessened because of rust. (iENEKAL REMARKS. Many inoculation, germination, and other physiological exi)eriments not mentioned in this bulletin but bearing closely on the cereal rust IDroblem, Avill be discussed in other publications of this Division. These experiments, like those made by Eriksson, show that in addition to the forms herein described, the cereal rusts have also distinct si)ecial- ized forms on various grasses, but as these forms do not occur on the cereals they are usually of no econonaic importance. In addition to the experiments already reported in a former publication (15, pp. 448-452), numerous others were made b}^ the writer to determine the behav- ior of the uredospores from a chemical standpoint. Also hundreds of inoculations on cereals with the iiicidia of various plants other than barberry have been made for the purpose of discovering, if possible, other a'cidial hosts for the cereal rusts, but so far none have been found. Again, the morphological differences between the different forms of the cereal rusts are, in a few cases at least, of considerable imiwrtance, but it is thought best, under the circumstances, to discuss them in another publication. Something should be said concerning the value of the inoculation experiments reported in this bulletin. It may be claimed by some that the experiments which proved successful in the greenhouse and on young plants would probably result negatively in a greater number of cases if made out of doors on more mature plants. This is not only possible, but, as already shown, the writer's results make it very i)rob- able; and thereby the experiments are all the more valuable, as in the greenhouse they are more exclusive in nature and in no case is an infection likely to take place out of doors if it can not be x)roduced in 68 CEREAL RUSTS OF THE UNITED STATES. the greenhouse. In case there is an error, therefore, it is on the safe side, and the number of possible host plants is gTaduall\' more and more limited, until finally, by reverse infections, the actual host plants are established. As already stated, a number of instances of repeated experiments on older plants resulted negatively, although they had been successful in the first experiments. The ''mycoplasmatic'' theory, advanced by Eriksson (25, pp. 475-477) to exi)lain the manner of perpetuation of Pnccinia ghimarum from one season to another, has alreadj' been mentioned. A description of Eriksson's experiments with this rust seems to reveal conditions differ- ent from those existing in the case of any other cereal rust. Barley plants grown from the seed and completely shut oft" from external influ- ences showed rust aftei a time. The seed was sown in sterilized soil and grown in glass houses, into which air could pass only through cotton- wool filters. From other statements and illustrations in Die Getreide- roste. it seems probable that we have here-at last an instance of the piop- agation of grain rust through the medium of the germinating seed of the host, as in some of the smuts. Eriksson claims that, as he found no mycelium present, the rust exists in the form of a mycoplasma in the cells of the embryo in a latent symbiotic state during the interval between harvest and seeding time. As the rust in question does not exist in this country, the writer has had no opportunity to investigate the matter. 2so other cereal rust is known to live within the seed in any form or manner.^ SUMMARY. (1) At least six and probably seven distinct rusts affect the cereals of the United States. They are as follows : Orange leaf rust of wheat {Puccinia nihigo-vera tritici), orange leaf rust of rye (P. rnhigo-vera secalis), crown rust of oats (P. coronata Corda), black stem rust of wheat and barley (P. graminis tritici Eriks. and Henn.), black stem rust of rye (P. graminis secalis Eriks. and Henn.), black stem rust of oats (P. graminis avence Eriks. and Henn.), and maize rust (P. sorghi Schw.), (2) Though all these rusts except that on maize i>robably cause some injury occasionally, the black stem rusts of wheat and of oats are by far the most destructive. (3) All tlie rusts are pretty evenly distributed over the United States, wherever their respective hosts are grown, except the stem rusts, which are perhaps most prevalent in the region between the Alle- ghanies and the ninety-fifth degree of west longitude north of the ' Cases where rusts other than those of cereals live over in the seed are not, how- ever, unknown. The actual peridiaof the ivciuium of Uromijces euphorhicp C. and P. may often be seen within the seed of Exiphorhia dentata. The writer's investigations concerning this rust are yet to he published. In 1890 Ralph (61. p. 18) also described an ;ccidiiim affecting the see,cidium of Bhamnua lavccolata infects oats, PhaJaris caroliniana, and Arrhenatherum elaiius. 2 Since this bulletin was prepared Dr. H. Klebahn has published in Zeitschr. fiir Pflanzenkrank., Feb. 4,' 1899, Bd. VI?I, Heft 6, pp. 321-342, a lengthy article entitled "Ein Beitrag zur Getreiderostfrage." It consists chiefly in an account of experi- ments made with the different grain rusts, aud of the results, which seem to conflict more or less with those obtained by Eriksson. In Centralbl. f. Bakt., Parasit., u. Infekt., Dec. 19, 30, and 31, 1898, Abt. II, Bd. IV, pp. 855-859, 887-896, 913-910 (6 illus.) appears an article by Prof. H. L. Bolley, entitled "Einige Bemerkungen liber die symbiotische Mykoplasmatheorie bei dem Getreiderost," in which the author, on the b.asis of his own experiments, also criticises Eriksson's theory, and gives the results of other iuvestigatious, which show the great vitality of uredospores and ajcidiospores aud the possibility of uredospore germ tubes passing through the actual epidermis of the leaf as well as the stomata. The original of this paper (in English) was presented before the Boston meeting of the American Association for the Advancement of Science, August, 1898. BIBLIOGRAPHY. 71' 9. BoLi,RY, II. L. Note ou the Wheat Rust. Am. Mou. Micr. .lour., 1890, Vol. XI, pp. 59, fiO. 10. BOLLEY, H. L. Wheat Rust: Is the Infection Local or General in Orijriu? A * r .■'" "^ ,<■ a t D.CiPa8smore;ael. ^-z. ■ •,■, - . ,305ton (hmparutive Resistaficeoffourva7ieties ofvjheatfram Oj^ancjie Leal' liust. BuL. 16, Div.Veg. Phys. and Path , U.S. Deft. Agr. Plate II. It A ' ■•'i BwtftiiBirtiaUsa, Xz- and Theiss Wheats Rusted in 1otato or in nutrient agar (ri. I, 0), producing a white, immediately much branched, many septate mycelium which by interweaving soon conceals the ascospore and in a short time (cowpea fungus) reproduces ])erithecia, first, how- ever, and usually within thirty-six hours, producing from the ends of short branches or from terminal hyphae (PI. I, 7) numerous, colorless, oval to narrowly elliptical conidia which are mostly 8 to 10 by 3 to 6 ^, straight or curved, continuous (or occasionally one-septate after falling off) and indistinguishable from those of the internal fungus (see below). Mycelium on slightly alkaline meat infusion peptone agar at first pure white, then grayish white (stromatic), producing immature but nearly full-grown peritiiecia in largo numbers within a week of the sowing of the ascosi)ore. Mycelium on steamed potato cylinders in test tubes at ' Altogether 244 ascospores have been ineasiired directly from the plant, and all of the sizes observed are recorded below in niicroua, together with the plants from which taken and the nnmber of each size. P = Cowpea; AV= Watermelon; C^ Cotton : 18 X 12 W ; 17.7 x 10.2 P' ; 16 x 15 ?■ ; IG x 14.5 P' ; 15 x 12 P' ; 15 x 11.5 W ; 15 x 10 W ; 14.7 X 12 P' ; 14 x 14 P- ; 14 x 13.3 P' ; It x 13 P' C ; 14 x 12 P^ W C ; 14 x 11 P' W ; 14 X 10 P^ W^; 14 x 9 P' W-; 13.5 x 12 P- ; 13 x 13 P' ; 13 x 12 P' W^ C ; 13 x 11 P' W; ISxlOP' W; 13x9 W^; 12.7 x 12 P' ; 12.5 x lOP'; 12x 12 P^s W'^ C^; 12x11 P^W'; 12x10 PSW'C^; 12x9.5C'; 12x9W'; 12x8 P' W"; 11.5 x 11 W ; 11. 5x 10.5 W ; 11.5 X 10 C ; 11.5 X 8 W ; 11 x 11 P- W* C ; 11 x 10 P2 W C«; 11x9 W C ; 11 x 8 W^; 10.5 X 10 P' C ; 10 X 10 P« W'"C--'; 10x9.5 C^; lOxOP'W^C^; 10x8W'; 9.5x9.5 W C-; 9x9 W-C^ 8x8C'. On the cotton most of the spores are globose, or nearly so, nearly all have a wrinkled exospore, and the medium and smaller sizes prevail. On the cowpea the majority of the spores are globose and wrinkled, and the larger sizes prevail. On the watermelon a majority of the spores are elliptical, many are smooth and the smaller sizes are the more common, as may be seen from the preceding figures. These variations do not, however, appear to have any specific value, since the spores derived from a single ascospore sorted out and sown on culture media may vary as greatly as any here recorded. This is shown by the following set of figures derived from measurements of ascospores taken from numerous perithecia grown on potato. The pure culture which was used for this purpose was derived from a single wrinkled ascospore taken from a typical perithecium on the cowpea. It will be seen that the general tendency of these spores is to exceed the size of the original spore rather than to fall short of it. They are also larger than the average of those taken directly from the host plants. Size of 180 ascospores from a luxuriant growth on steamed potato; original spore 12 X 12 Jit, the number of each size is given at the right hand above ; all with wrinkled exospore except a very few of the small ones: 22 x 12' ; 20 x 15' ; 18 x 14' ; 18 x 12'; 18 X 11' ; 18 X 10.5' ; 18x10'; 17x12'; 17xll'; 17x10.5'; 17x10'; 16.5x12'; 16x15'; 16x14'; 16x13'; 16xl2«; 16x11.5'; 16x11^; 16x 10' ; 15.3x 12' ; 15 x 15' ; 15x13'; 15 X 12.5' ; 15 x 12- ; 15 x 11- ; 14.5 x 13.5' ; 14.5 x 12.5 ; 14.5 x 12' ; 14.5 x 10' ; 14.3 x 12.5' ; 14xl4«; 14x13.5'; 14x13'^; 14x12.5'; 14 x 11.5'* all from one ascns; 14x11'; 13.5x13.5^; 13.5x13"; 13.5x12.5'; 13x13"; 13x12''; 13x 10 ; 12.5x 12.5^; 12.5x12"; 12 5x10.5'; 12 x 12-"- 12 X 11.5' ; 12 x 11' ; 12 x 10' smooth (7 others from the same ascus were smooth); 11.7 x 11.7'; 11.5 x 11.5' smooth; 10 x 10- smooth. Thirteen of the elon- gated spores were ovate ; the rest were elliptical. A very few spores were flattened on the side 3 from the pressure of the ascus. 10 first suow-wbite, then developing a pnft'ed-np, grayisli white, tongh stroma which becomes thickly studded with bright coral-red perithecia in about eight days from the sowing of the ascospores. Steamed sterile potato has proved a very suitable medium for the growth of this stage of the fungus, the perithecia becoming ripe and discharging their spores copiously ill two to three weeks from the sowing of the ascospores. In a number of instances the cycle from ascospores round to ascospores (ripe and discharged from the perithecium) was as short a time as twelve days and twenty hours. These ascospores were all taken from peri- thecia found on stems of tlie cowpea. Perithecia have also been grown from 150 or more microconidia (PI. 1,7). They were isolated by the poured-plate method, the material being derived from an agar culture about two weeks old which was made from ascospores. The perithecia developed on five diflerent media, viz, agar, banana, onion, carrot, and potato. On the first three media they appeared sparingly; on the carrot and potato they were abundant and appeared in about two weeks. The ripe ascospores were either shot out of the perithecia against the walls of the test tubes in such numbers as to make distinct brown patches or else were slowly extruded, crowning the ostiolum with irregular brown balls and clumps of spores (PI. 1, 1 a). So far as known to the writer, this is the second time perithecia have been derived from the couidial fructification of any Hyi)Ocreaceous fun- gus. Brefeld and von Tavel do not record any such case. On the con- trary, only three such cases are mentioned by them for the whole group of Ascomycetous fungi.' Klebs has since recorded this for Eurotium repens, and Hugo Gliick, in 1895 (Hedwigia, p. 254), reported it for Nectria moschata, which he obtained in about four weeks from pure cul- tures of the sickle-shaped conidia of Fusarium aqueductum, in water and plum decoction to which oak wood and bark had been added. A'ery likely Brefeld's numerous failures are to be attributed to the fact that he used his "Ntihrlosung" too exclusively, i. e., did not vary his culture media widely enough, so as to more nearly imitate natural conditions. On the contrary, the conidial stage of the watermelon fungus (spore taken in July from the interior of a vessel) has been cultivated for five years on a great variety of media, including potato, without showing a ' "As many yeast conidia remain under cultivation through endless generations always the same, so ordinarily the spores from couidiophores yield, in pure cultures, always the same conidiophore ; those of pycnidia always the same pycnidia. In like manner many oidia, under certain conditions, never produce anything else than oidia. Only in rare cases has it been possible to obtain the ascus fructification from the spores of an accessory fruit form. Of this the Penicillium [crustaceum] spoken of in the second Heft of this work remains the most interesting example. Addi- tional examples are Endomyces MaynusH and Diaporthe controversa. "Up to the present time little is known of conditions governing the development of the ascus fructification. Observations in nature frequently show its dependence on a certain duration of development, on the time of year, and on a definite sub- stratum."— Brefeld and von Tavel: Untersuchungeu, Heft 10, p. 349. 11 trace of peritliecia, altbougli from time to time special ettbrts were made to liiul a substratum which wouUl lead to the production of i)erithecia. This is the strain of fungus which has proved so actively parasitic in the hands of the writer. On agar and on potato the peritheciaof the cowpea fungus showed a distinct tendency to be larger than on the host plant, owing probably to extra good nutrition. The diameter of the largest on nutrient agar ranged from 320 to 376 //. The diameter of the largest on potato ranged from 3G0 to 410 //. The perithecia on the potato were exceedingly numerous, sometimes as many as 5,000 developing on an urea of not more than 15 or 20 square centimeters. They were much less numer- ous on the slightly alkaline meat infusion peptone agar, and ripened and developed their bright red color much more slowly. On certain other media the peridium remained nearly colorless or the perithecia failed altogether to develop, although the fungus was under the same conditions except as to the substratum. Numerous cultures on a great variety of media have shown that the presence or absence of a stroma is entirely a matter of the substratum. :So attempts have been made to grow the ascospores from the similar I)erithecia which have been discovered on the dead stems of cotton and watermelon. No perithecia ever developed in any of the cultures made from internal or external conidia taken from the cotton or watermelon. CONIDIAI. FRUITS. (1) Microconidin. (Cephalosporium stage).— Colorless, oval to nar- rowly elliptical, straight or slightly curved, non-septate spores, 4 to 25 by 2 to //, borne singly one after another (PI. Ill, 3) on the ends of short branches of a mycelium, which fills the water ducts and interior parts of the liviny stem (melon, cowpea) with a dense growth that is pure white when seen in mass in the stem or when cultivated out on alkaline meat infusion peptone agar, or on other alkaline media, even rice: conidia frequently 1 septate (or rarely 2-septate) after abscission, in cultures often remaining in a little group n round the end of the stationary couidiophore or becoming scattered when the latter elongates after having produced a group of spores, as is frequently the case (PI. 1, 7, PI. Ill, i), aerial or submerged {Fusarimn vasinfeetum Geo. F. Atkinson, described from cotton and okra; Fvsarium nivenm'Erw.F. Smith, described from watermelon). None of the large, lunulate, 3 to 5 septate spores describ( d below have ever been observed in the vessels of any of the host plants or submerged in fluids or solids. (2) Macroconidia (Fusarium stage).— Lunulate; 3 to 5 septate spores, 30 to 50 by 4 to // (PI. I, 9; PI. II, 7; PI. V, 5), borne on the surface of dead stems in immense numbers on innumerable small, oval or hemispherical conidia beds which arise from the internal mycelium and consist of compact, irregularly branched, short conidiophores. Examined in water the single spores are nearly colorless; in mass they 12 are at first wliite, but very soon after the formation of the eonidia beds they become colored, the variable tint raDging from pink, pale tlesh color or pale salmon to deep salmon. When germinated in water or under acid or alkaline agar or iu very moist air, producing eonidia indistingaishable from those borne by the internal fungus (PI. II, 10; PL 111,11 in part).' Immature spores from the eonidia beds are, of course, much shorter than the measurements given above, and for a time are non-septate or only 1-septate. When accidentally knocked off, such spores appear to be still capable of growth and septation. An effort was made to identify this external conidium with some one of the many described forms of Fusarium, but without success. The above measurements are derived from hundreds of spores and are believed to give nearly the limits of variability, but it would be quite possible to make half a dozen species of Fusaria from the material I have had in hand, if only a few measurements were made and a hasty description written out, as is frequently done. Indeed, there is no doubt that the literature of systematic mycology contains much of such rubbish. One has only to study critically a few members of this group and then turn to the descriptions iu Saccardo's Sylloge Fungorum to be convinced of it. Not only is the identification of Fusaria from descriptions usu- ally impossible, but very often with the specimens themselves the case is little better, owing to the fact that many Hypocreaceous fungi bear eonidia which have been put .into the form-genus Fusarium and which so closely resemble each other (see the plates in Vol. Ill of Tulasne's Carpologia, and iu Heft 10 of Brefeld's Untersuchungen) that mor- phology alone is of very little assistance in their identification. In many cases, at least, the only safe way is to make cultures and inocu- lations if the fungus is a well-marked parasite ; or to obtain the less 1 Old cultures of the iuternal watermelon fungus, when transferred from sterile horse dung to potato broth, produced, along with the small elliptical non-septate eonidia, all sizes and shapes up to those which were luunlate, 3 to 4 septate, and 38 M long, hut none 50/1 long could be obtained in this medium, nor were any of them distinctly salmon colored. These cultures were all derived originally from a single non-septate microconidium separated out by the poured-plate method. When hothouse watermelons were grown iu soil infected with pure cultures of the internal melon fungus the macroconidial stage frequently appeared on the sur- face after the wilting and death of the ])lants. One of these big, lunulate, 3-septate, external eonidia was separated from its fellows and cultivated for a long time, first in sterile horse dung and subsequently in potato broth, in the same manner as the preceding. The mycelium derived from this spore produced great numbers of micro- conidia, indistinguishable from Ihose borne on the internal mycelium. At least three-fourths of the whole number of spores were of this type, but there were also plenty of typical macroconidia, spores 40 to 50 /i by 4 to 6 /i, lunulate, and 3 to 5 septate. Betweeu these two extremes there were all possible gradations, showing clearly that little dependence can be placed on statements respecting shape and size of spores in the ordinary systematic descriptions of Fusaria. Color is likewise misleading, since the same eonidia bed may be white, pinkish, pale salmon, deep salmon, etc., according to its age and the character of the substratum on which it has developed. 13 variable peritheeia from pure cultures of the eoiiiilia, sometliiug vre have uot yet learned to do with any certainty; or, linally, to discover in nature the j^erfect fruit form and derive the other I'oinus from it. So far as known to the writer, no Fusaria except those above mentioned have been described from cotton, watermelon, or cowpea; but even if they had, it would in the present stage of our knowledge be next to impossible to establish identity beyond reasonable doubt, since many species of Fusarium are believed to be purely saproi)hytic and are known to grow on almost any dead substance, and one such form has been discovered by the writer in Washington on a dead stem of cow- pea in connection with the peritheeia of a Nectria. (3) Chhimydospores. — On the surface of the dead stems of the water- melon and in old cultures of the melon fungus on horse dung, globose, thin-walled, smooth, terminal or intercalary bodies ai)pear, and in mass on the dung are brick red (I'l. Ill, 12). These are part of the life cycle of the fungus and nppear to be chlamydospores comi)arable with those of Hyponn/res solani described by Keinke and Berthold.' They api)ear to have an outer and inner wall and germinate readily in water, but have not been studied as critically as the other spore forms. They were found in many test tube cultures, and have been observed a num- ber of times on the surface of the dead stems associated with the macro- conidia. They are usually 10 to 12 /.i in diameter, the extreme limits of those measured ranging from 7 to 15 ii. (4) Pyc7iidia{.^).—^o pycnidia have been seen either on the host plants or in any of several hundred cultures made on a great variety of media. The cultures began in the summer of 1894 and are still in progress. The examinations on the host plants have included hundreds of dying and dead specimens collected in diflerent years and in various locali- ties from July to October. From these observations the writer believes he is warranted in concluding that there are no pycnidia in the life cycle of this fungus. EFFECT OF MODIFICATIONS OF THE SUBSTRATUM. The effect of modifications of the substratum on (1) the production of a stroma, (2) the color of the mycelium, (3) the color of the perithe- eia, and (4) the development of the peritheeia, is described ill the following i^aragraphs: (1) The production of a stroma. — The insignificant subiculum on the host plants and the thick stroma on potato have already been con- trasted. Tlie limits of the stroma on potato are beautifully differen- tiated when thin cross sections are put into chlor-iodid of zinc. The starch-bearing cells of the substratum become blue, while the stroma becomes yellow, and the exact limits of fungus and substratum are easily distinguished. On ordinary nutrient agar the cowpea fungus produces only a feeble stroma. The addition of 1 per cent cane sugar ' Die Zersetzung der Kartoffel durcli Pilze. 14 to such agar increased the tbiekuess of tbe stroma. The stroma was better developed than in the same agar with only 0.2 per cent sugar. In agar with 12 per cent cane sugar there was a great increase of this tissue, which formed a sort of monstrous stromatic mountain the whole length of the long slant. This ridge was 1 centimeter wide and fully one-half centimeter high in its highest parts. Its very irregular, tough, gray- white surface was covered with a thin layer of white aerial hyplne, and 24 days from sowing was thickly set with colorless perithecia. The latter were, however, also visible at this time on the scanty, non-stromatic mycelium which had climbed up on the walls of the tube above the agar. The formation of perithecia in a similar way was also observed in other culture media. So far, therefore, as this fungus affords any basis for judgment, the stroma is one of the most easily variable parts of a fungus; consequently it appears hazardous to make separation of closely related forms into distinct species solely on the presence or absence of a stroma. (2) The color of the mycelium,— The variability in color according to the nature of the substratum has been a subject of much interest to the writer. In the interior of the host plants the fungus is usually pure white, except in cotton, where the older mycelium is frequently brown. For a considerable time the writer supposed pure white was its only color, but trial of a variety of culture media brought to light some very unexpected and most astonishing facts. The melon fungus is pure white on slightly alkaline nutrient agar. The fungus remained snow white for over a month on alkaline corn starch steamed in distilled water with asparagin added. In another series it was pure white for 11 days, but had become faint rose color on the fourteenth day and remained so up to the twenty-fifth day. On boiled rice, with the addition of bicarbonate of soda, the fungus remained pure white for 40 days. On crushed cowpeas, steamed in distilled water, tht^^ fungus made a copious growth from the start, but remained snow white as long as the cultures were under observation (10, 17, and 22 days). The fungus was pure white in slightly alkaline, peptonized beef broth, and did not become colored on addition of cane sugar or of dextrin. On acid, neutral, and alkaline nutrient gelatin the fungus madea pure white growth. In acid potato broth the fungus was white for 11 days, except for certain hyphal strands which had climbed out of the fluid and were attached to the walls of the tubes. The submerged growth and the pellicle were pure white. The following acids were tried with this broth : Malic, oxalic, succinic, tartaric, and citric, 1 c. c. of the -i^ normal solu- tion being added in each case to 10 c. c. of the broth. The behavior of the fungus was the same in each case, white on and in the fluid, with brown hyphte on the walls above it. On strongly alkaline litmus agar the fungus was white on the sixth U" o day. During the first three days it had made a slow, poor growth, but it now covered the surface with a matted growth from wliicli no hyjjhte projected into the air. This medium consisted of JO c. c. agar, 20 drops of saturated sohition of sodium carbonate, and 2 c. c. of Shai-p & Dohme's violet litnuis solution. On the fouiteenth day there was no reddening of the substratum and the surface was covered with a white mycelium, but some of the hyphii- tlireads which projected out into the air were bluish. At the end of 112 days the agar was nearly dry. It was still blue and was covered by a bluish-white film of fungus. On the glass above the agar many of the filaments of the fungus were now bright blue. This stain most ])robabl3^ came from absorbed litmus rather than from any pigment manufactured by the fungus. In horse dung sterilized with distilled water the fungus made a'white growth (some yellowish hyphie). On banana the fungus was white on the start, but in old cultures it was rose, salmon, or purple, and patches of it became bright orange buff. In 12 days the fungus made a copious growth in the acid juice of Concord grapes, and its color varied from snow white or a very feeble tint (above) to purplish and rose color (below). For some days the fungus was barely able to maintain itself in this very acid medium, and was white, but at the end of 7 days the compact islands of mycelium were roseate. After 42 days the fluid was still strongly acid, and the red color persisted on litmus paper for some years. After -1 months the prevailing tints were white to reddish brown. The very numerous mycelial strands on the walls of the tubes above the medium Avere a fine red brown (between burnt sienna and tawny). On Spanish onion cooked in distilled water the fungus made in 3 days a moderate growth, and its color varied from white to lilac (next substratum); in 5 days the fungus was purplish; in 8 days there was a distinct increase of purple and rose color, but there was no color in the substratum under the water. After 21 days the lower part of the onion iu both tubes (that part under water) retained its natural yellow color, although a microscopic examination showed it to be grown full of the hyphfe of the fungus. At this time the aerial mycelium was white, rose, purplish, salmon, and brown (a few strands on the walls of the tube). At no. time did any crimson color appear. In 10 c. c. of Dunham's solution with 1 per cent malic acid the fungus filled the fluid and became bright rose color in 20 days, but the acqui- sition of the color was slow. The litmus reaction was still decidedly acid on the twentieth, thirty-fifth, and forty-fifth days. This experi- ment was repeated. In 13 days the mycelium was purplish. The addition of 2 per cent cane sugar to this acid Dunham favored vege- tative growth, but did not increase the production of color or hasten it. On the fourteenth day in one tube and on the fifteenth in the other the fungus was still pure white. This experiment was repeated. 16 Sugar did not help ou the formatiou of color, but rather seemed to retard it. The litmus reaction was decidedly acid ou the twentieth and twenty-seventh days. There was marked retardation of growth in Dunham's solution with 2 per cent cane sugar and 1 per cent sodium carbonate. After 20 days the fungus still seemed white, but the fluid was now slightly yellowish. There was less growth than in the acid Dunham. In Dunham's solution containing 2 per cent cane sugar and 0.4 per cent sodium carbonate there was a much better growth on the start than in the same medium with 0.4 i^er cent of malic acid substituted for the soda. At the end of 13 days the fungus filled the alkaline saccharine fluid and was yellowish brown; the fluid was also browned. The brown color in the mycelium was noticed as early as the eighth day. The extreme upper part of the gelatin-tube cultures (the substratum) was finally stained rose brown. In test-tube cultures 1 to 10 of November 27, 1894, on bread steamed in distilled water (5 consecutive days), the fungus grew luxuriantly. On December 18 the bread was nearly or quite covered and hidden by the fungus. In the top of the cultures where it projected into the air the fungus was snow-white. Farther down this ])ure white color shaded into yellows, purples, flesh tints, and reds, the predominating color being an irregularly distributed bright crimson. All of the tubes showed this crimson color, some more than others; several were very striking. There was less purple and more crimson than in the same fungus when grown on potato. In days, on pearl tapioca steamed in distilled water, the fungus was purple. In 3 days, on crushed wheat steamed in distilled water, there was a copious growth and a bright purple seam where the fungus rested on the substratum. In 6 days from the inoculation the purple color had extended and a brilliant crimson was visible in the hyphte bordering on that part of the substratum first attacked. The fungus resting on the wheat was purple, while the upper parts of the same mycelium were pure white. In 3 days, on hominy steamed in distilled water, the fungus made a- copious growth and developed an abundance of color. This was pur- ple in the greater portion, but crimson in the oldest part. On the sixth day the purple color involved the upper half of the substratum in one tube and nearly the whole of it in the other tube. The rice cultures proved particularly interesting. On the third day, in tubes of rice steamed in distilled water, there was a copious devel- opment of mycelium, which was snow-white in the air and i)urplish and crimson on the substratum. In 5 days the bulk of the culture was purplish to carmine. Even the aeriel mycelium was tinged. A little of the mycelium in the bottom of the tube and some on the walls of the tube above the rice was still colorless. After 20 days this crimson 17 rice was stirred up in distilled water which became pink even when 20 to 30 c. c. was used with the contents of but one test-tube culture. This fluid was strongly acid to litmus. On tilteriug it through tissue paper it was colorless. On boiling it gave oli' a pleasant aromatic odor and litmus paper held in the steam was reddened. Most of the acidity passed away on boiling. The color of the uufiltered fluid changed from pink to purple on boiling in a sterile test tube. There was no change of color with 1 per cent chromic acid. Six rice cultures were now instituted in test tubes as follows, each, with one exception, holding about the same quantity of tlie cooked rice: (1) llice with 1' drops of saturated solution of sodium carbonate; (2) rice with 3 drops of the soda solution; (3) rice with 5 drops of one- half per cent solution of hydrochloric acid; (4) rice with 10 drops of this acid water; (5) rice made blue with 2 c. c. of violet litmus solution; (G) rice made blue by 2 c. c. of violet litmus solution and 2 drops of saturated solution of sodium carbonate. These tubes were all inocu- lated in the same way, from one of the crimson rice tubes just described, and were under the same cultnral conditions. They will be described by number. In 2 days most of the mycelium in 1 was snow-white, but there was a slight purpling around two rice grains; in 2, which diflered from the others in having only about two-thirds as much rice and was, therefore, nearly twice as alkaline as 1, the fungus was growing rather slower and was i»ure white throughout; in 3 there was a decided pur- pling of most of the fungus and of the rice grains which were touched by it; 4 was like 3; in 5, that part of the fungus projecting into the air was pure white, but the blue rice grains in the vicinity of the growing fungus had become bright purple; in G there was also some purpling of the rice grains attacked, but it was more restricted, and, as in 5, the bulk of the rice was bright blue. On the 6th day the condition was as follows: 1, fungus has made a good growth and some of it is pale purple to light rose, but there is not one-tenth as much color as in the tubes which received the acid; 2, the fungus has grown much within the last few days, but there has been no corresponding increase of the color, which is now a faint, scarcely noticeable rose or purple, i. e., there is not one-hundredth j)art as much color as in the acid tubes; 3 and 4, very brilliant and beautifal, the prevailing color, crimson, shading into rose and purple, but where the fungus x)rojects above the culture into the air it is pure white, or shows only the faintest tinge of pale rose; 5, the upper one-half of the tube is tilled with a copious growth of the fungus, and most of the blue color has disappeared, being- changed below into violet and above into rose color, and brilliant purples and crimsons, but that part of the fungus which grows into the air above the culture is white to faint rose; G, like 5, except that tne lower one-half of the rice is still a pure blue. On the 9th day tubes 1 and 2 oifered a striking contrast to 3 and 4. In tube 1 the fungus was becoming quite roseate over large areas, but none was crimson or 4133— Ko. 17 2 18 purple; in tube 2 (which received twice as much alkali) the fungus had grown through all parts of the rice, and varied from snow white to a faint roseate color, which was scarcely noticeable. In 3 and 4 the starch grains were covered and hidden by the fungus, which was snowy only where it projected into the air above the culture. There were no purple hues, but the body of the culture varied from rose to crimson, the latter color jiredominatiug. There was now most crim- son in the tube which received the most acid, but possibly this was accidental. In 5, all of the blue had disappeared, but the rice grains at the extreme bottom of the tube still showed a very little red j)urple. All the rest of the substratum was so closely invested by the fungus that the rice grains were hidden. The color of the fungus varied from rose color to the most vivid crimson, the latter color predominating, the only white part of the fungus being that which projected into the air at the top of the culture. In 6, which was more alkaline, a half cubic centimeter of rice in the bottom was still violet color, none was blue, and in other respects the culture was like 5. After 40 days 1 and 2 still showed the restraining eflect of the soda. In 1, a little of the mycelium was now crimson, but most of it was only pale rose color; in 2, there was much less color than in 1, most of the color being pale rose, but with a little violet and a trace of crimson; the rice had shrunken from one side of this tube and the cavity was filled with white mycelium. In 3 and 4, crimson was still the i)revailing color; both were shrunken away from the wall on one side, and in one tube white mycelium was beginning to fill the cavity. In 6, the lower one- third of the rice was now brilliant crimson, and the upper two-thirds, which had lost much water, was changed to dark purple and partly overgrown with white mycelium. The dark purple color and the crim- son could be seen distinctly in the hyphal strands as well as in the rice grains themselves. In 6, at the end of 58 days, the entire mass was shades of violet, no crimson color remained, and the secondary super- ficial mycelium was white. The above experiment was repeated as follows with five test tubes of rice boiled in distilled water : (1 ) Rice, with 3 drops of saturated solution of sodium carbonate; (2) rice, with 6 drops of the soda solution; (3) rice; (4) rice, with 5 drops of one-half per cent hydrochloric acid; (5) rice, with 10 drops of one-half per cent hydrochloric acid. All were inoculated at the same time in the same way and from the same culture. In 6 days there was a large amount of color in tubes 3, 4, and 5 — rose, purple, and crimson. The color was as well developed in the pure rice as in the tubes which received the acid. Tubes 1 and 2 behaved much as in the preceding series. In 1, at the end of 14 days, the entire rice was invaded, and some of the mycelium was roseate to faint purplish, but there was not one-hundredth part as much color as in the acid tubes. In 2, the whole substratum was invaded, and the mycelium was white to faint roseate, there being about one-fourth as much color as in 1. On the twenty-fifth day, in 1, the color was much deeper than on the four- 19 teenth, but it was still mostly white or roseate, there being very little crimson. lu 2, the fungus was white to rose color, with occasional hyphal strands becoming crimson. The color had increased, but it was only about one-half as much as in 1, In 3, the body of the culture was a beautiful crimson mottled with purple. The color was lodged both in the substratum and iu the hyphal strands. Where the rice had shrunken away from the wall a net work of white mycelium was growing out of the crimson mycelium and covering it. In the upper part of the tube, where the fungus projected into the air, it was also white, or rather, so nearly so that it appeared tinted only when taken out and massed on white paper. The conditions in 4 and 5 were like those iu 3, except that there was no i)urple color. In both tubes the rice was shrunken away from the wall and white mycelium was lilling the cavities. This contrasted curiously with the crimson hyphai and rice grains from which it was growing. A tube of litmus rice made much more strongly alkaline than the two j)reviously mentioned was inoculated October 28, 1895. ( )n Novem- ber 14 the upper one-third of the rice (that best aerated) was thoroughly invaded by the fungus, the deeper parts being free, or nearly so. The fungus was snowy white. The rice grains invested by the fungus were changed from bright blue to a pale blue, but there was no purple or red color. This culture consisted of 15 c. c. cooked rice, 2 c. c. of violet litmus solution, and three-fourths cubic centimeter of a saturated solu- tion of sodium carbonate. On December 5 the rice had lost all of its bright blue color, but there was no violet or crimson. The entire cylin- der of rice was more or less overgrown and interwoven with the fungus, which was white; the rice grains in the upper part of the tube were greenish gray, and in the lower part a pale purplish rose. In cutting through old dense conglomerated rice cylinders overgrown with the crimson fungus, the periphery was generally crimson, while the center of the cylinder was mottled, the exterior of the rice grains being red and their interior white. This seems to be another indica- tion that the pigment requires an abundance of free oxygen for its development. Three large Erlenmeyer tiasks of rice were prepared as follows: (1) One-half pound of rice sterilized in several hundred cubic centimeters of distilled water; (2) one-half pound of rice boiled in several hundred cubic centimeters of distilled water, then knocked out of the flask and intimately mixed with one-fourth pound of c. p. calcium carbonate, and subsequently resterilized with more water; (3) one-half pound of rice boiled in several hundred cubic centimeters of distilled water, then 100 cc. of a saturated solution of sodium carbonate added, and the sterili- zation completed. These flasks were inoculated at the same time, in the same way, and from the same culture. In 48 hours the condition was as follows : In 1, the fungus had begun to make a vigorous growth from four dif- ferent centers of infection; color, i)urplish for the most part, but some 20 crimson. Two rice grains wliicli projected up above the general sur- face, and which were invested by the fungus, were wholly deep purple verging to crimson, while the hypha; projecting from them were color- less (white). In 2, nearly as much growth as in 1, but much less color, the hyphie being white above and purplish where they rested on the colored sub- stratum, but not crimson. In 3, fungus pine white, but only a trace of growth; i. e., not one two-hundredth as much as in 1 or 2. At the end of 13 days the condition was as follows: In 1 the surface was nearly covered with beautiful masses of crimson mycelium; the upper half of the rice was also invaded by the fungus and was mottled — white, purple, and crimson ; 2, too much water in this Hask, and fungus growth conspicuous only on the surface; no crimson stain, but some pur- ple hypha^; nearly all of the fungus was snow white; in 3, fungus snow white ; there were now five small tufts of fungus on the surface of the rice, but the latter was too alkaline for a rapid growth. In 21 days the condi- tion of l!^o. 3 had much improved, the fungus being visible on a hundred or more of the surface grains of rice as a thin white hyphal layer. On the thirty-seventh day the fungus had increased noticeably in growth, and all of it was pure white, and the rice grains were also free from color. That the alkali did not simply mask the color was shown by the fact that when some of the fungus-infested rice grains from this flask were put into dilute hydrochloric acid there was no change of color. In flask 1 the fungus had not reached the bottom of the rice at the end ot 21 days, but there were traces of it nearly to the bottom. The upper half of the culture (a large mass, it will be remembered) now varied from rose color to crimson, both the rice and the fungus, and was very showy. After about 40 days, during which the bright color increased in some parts and faded to purple in others, the top part of flask 1 was extracted with hot alcohol. The alcohol became brown-red and a bluish-violet residue remained. This brown-red alcohol became opalescent on the addition of water, but the precipitated color passed readily through a filter paper. The alcoholic extract faded slowly on exposure to light. On another occasion some of the bright-red fungus was extracted with alcohol. In quantity the alcoholic extract was dragon's-blood red, and the residuum was blue-purple. The red alcoholic extract was rendered colorless by a few drops of strong caustic-potash liquor or caustic-soda liquor, but was not destroyed by liquor-ammonia fort., even when used in large quantities. The addition of a saturated solution of sodium carbonate did not destroy the color. It would seem, there- fore, that two colors may be present— a blue and a red — the latter solu- ble in alcohol and easily destroyed by light and by caustic potassa and soda, the former insoluble in alcohol, more resistant to light and unaffected by alkalies. The red fungus from a rice culture was tested with various reagents 21 and the results are liere given for comparison with those obtained with the red color of the perithecia. which are given below: Strong am- monia, exposure to the Aapor, even for an iustaiit, changes the red to purple; fragment still purple after 42 hours in the liquor; fluid itself slightly tinged. Ca untie- soda liquor, red changed instantly to very dark purple, almost black; purple after 42 hours; fluid slightly tinged. CauHiivpotash liquor (strong), red changed immediately to purple; after 42 hours still purple; lower one-third of the licpiid reddish — i. e., much more decidedly tinged than any other alkaline fluid. Sodium carbonate (saturated solution), red changed at once to purple; purple after 42 hours, and the fluid no longer ([uite colorless. Ammonium carbonate (saturated solution), red changed at once to purple — it changes even in the vapor; purple after 42 hours; fluid as in the preceding — i. e., a very slight departure from colorless. Chloral hydrat (saturated solu- tion), no immediate change unless to become brighter red ; soon soluble; after 42 hours fluid still roseate; fungus dull red. Sulphuric acid (1 l)er cent), no change in -12 hours; fluid colorless, fungus red. The same result was obtained with 10 per cent sulphuric acid. Nitric acid (1 per cent), no immediate change, and none after 42 hours; fungus still red and fluid colorless. Hydrochloric acid (1 per cent), no change in 42 hours; fungus red, fluid colorless. Acetic acid (1 per cent), no change in 42 hours; fungus still red and fluid colorless. ^Vood alcohol, dis- tinctly soluble; alcohol around the red fragments of fungus colored within a few minutes; after 42 hours fungus purple, fluid pale wine-red. Ethyl alcohol (common, 95 per cent), color soluble; the alcohol around the fungus became colored in a few minutes; after 42 hours fungus purple, fluid a pale wine-red. Sulphuric ether, no immediate action, but after an hour or two the ether was tinged; after 42 hours fungus pur- plish, fluid pale brownish-red. Chloroform (purified), no immediate change, and none after H hours; after 42 hours fungus purplish, fluid pinkish-red. Benzine (purified), no change in 4.3 hours; fungus still red and fluid colorless. Benzole, no change in lA hours; after 43 hours fungus red to purple-red, and the fluid with a slight pinkish tinge. Carbon bisulphide, no change in 43 hours; fungus still red and fluid colorless. One test was made to determine whether the fungus would produce its colors on uncooked rice in water. In 2 days a distinct purple color appeared on the rice grains infested by the fungus. All of the above experiments relative to color production were made with pure cultures of the melon fungus. Allusion has already been made in several places to the growth of white mycelium from colored, and vice versa. It happens very fre- quently that on mycelium of the same size and apparently of the same age one branch will be colorless (white) and another deeply stained (purple, red, blue, brown, yellow). The staiued hyphte were usually quite granular, but unstained hyphae were also sometimes observed 22 with granular contents. Went observed the same thing in Monascus (Ann. des Sci. nat. Bot. 1895, No. 1), and the writer has also seen it in that fungus. The cowpea fungus also stains various starchy substrata, but the writer was never able to get as bright colors as with the melon fungus. The red color of the perithecia is insoluble or nearly so in ether, chloroform, benzine, benzole, or carbon bisulphide. The color is very slowly soluble in ethyl and methyl alcohol, but decidedly more so in the wood alcohol than in the 95 per cent ethyl. That is, after 48 hours there was no change in the ethyl alcohol, and even after several weeks most of the perithecia were still bright red, although a few had faded; after 48 hours in the methyl alcohol the red perithecia (from a potato culture) were a little paler, and after several weeks a few only were pale red, most of them being pure white. In a saturated solution of chloral hydrat the color changed at once slightly, was dull orange-red in 15 or 20 minutes, yellowish-red ia 2 hours, and dull yellow in 48 hours. In chlor-iodid of zinc the color soon changed to dark purplish red and then to brown. The mycelium and interior of the perithecia did not blue, but changed from white to yellow. In mineral acids (nitric, sulphuric, hydrochloric, chromic), strong or weak (weakest solution 1 per cent), the perithecia changed instantly from red to yel- low and remained yellow. The same change took place in acetic acid, but required several minutes when a 1 per cent solution was used. In a saturated solution of ammonium carbonate the perithecia were purple at the end of 48 hours, but there was no visible change during the first 2 hours. In strong ammonia water they changed immediately to pansy- purple, and did not at once lose color, but were paler at the end of 48 hours. A purplish stain has frequently been observed in the parenchyma of old dead stems of cotton, melon, and cowpea attacked and killed by this fungus, and this stain when subjected to certain solvents behaves in the same way as the red pigment of the perithecia. The formation of purple or other bright colors in the substratum, and occasionally in the mycelium itself, appears to be common to many Fusaria. Schacht observed it many years ago (Prings. Jahrb., Bd. Ill, pp. 446-447). In potatoes attacked by wet rot he found the cavities lined by Oidium violaceum Harting. This lining was dark violet to blue black. This color he states to be due to the contents of the mycelium and to the effect of the latter on the substratum. On addition of sul- phuric acid the mycelium and spores became rose red. This change, he thonght, denoted conversion of starch into sugar, for it took place eveu inside of starch grains, and the color is similar to that which may be obtained with sugar, albuminoids (Eiweissstoff), and sulphuric acid. In dry starch attacked by the fungus and kept since 1855 he could, however, get no reduction of the copper on warming in Fehling's solu- tion. Oidium violaceum appears to correspond to the chlamydospore 23 stage of ITeocosmospoia. Scbaelit states that it is only one stage of Fusisjwrium solani, lie having found both spore forms attached to the same mj^celiiim. K. Klein (Beitrag z. Kent. d. lothen Malzschimels, Mitt. d. osterr. Vers. Stat. f. Braiierei n. Miilzerei, VVien, V, 1892) has also found a Fusa- rium causing bright color reactions on starchy media, but 1 have not seen his paper. My conclusions relative to the formation of color by the melon fungus are as follows: (a) On neutral or acid media in the i^resence of free oxygen and of starchy foods — e. g., potato, bread, rice, tapioca, wheat, hominy, cucum- ber agar, etc. — this fungus develops in the substratuni a series of the most brilliant colors (I'l. 1, 13), which are then absorbed by the hyph*. These hues include many shades of pink, red, purple, and violet, and in some of the substrata— e. g., bread or boiled rice — are particularly brilliant, changing gradually from shades of purple and rose color into the deepest crimson (rose carthamine). This color is much brighter and purer than any 1 have been able to obtain with Went's Monascus purpurcus. During the development of this pigment the substratum becomes intensely acid (mostly OO2, but some lactic acid according to Mr. K. P. McElroy). If, however, alkaline substances (caustic lime, carbonate of soda, etc.) be added to the substratum in advance, so as to neutralize the acid or acids as fast as formed, no color is devel- oped, the fungus remaining snow white, as in the vessels of the melon plant. If less alkali be added, the colors appear gradually after a time, which is longer or shorter according to the amount added. {b) The yellow and brown colors are formed in the presence of an alkali, but apparently not unless sugar is also present. According to the writer's view, the brown stain of the lignified walls of the vessels is due beyond question to the presence of this fungus. Since the lignified walls are much more apt to be stained than the pure cellulose walls, it woidd seem as though the presence in the lignified wall of coniferin or of some related substance may have something to do with the produc- tion of the brown stain. The vascular bundles in melon, cotton, and cowpea contain a distinctly alkaline fluid and the fungus is able to dissolve its way through cellulose walls. If it should also be able to split up coniferin, or some similar glucoside, held in the walls of the vessels, with the liberation of a sugar, then a brown stain might per- haps be formed just as readily inside of the plant as in the alkaline peptone water to which cane sugar was added. "^ (3) The color of the peritliecia.—Eo white- walled or colorless perithe- cia have ever been observed on any of the host plants. All have been bright red. They have also been red on a variety of artificial media, particularly those on which they have grown best— e. g., steamed potato, crushed wheat, malic acid agar. Red must therefore be assumed to be their natural color. 24 The color of the peridium is, however, much more easily modified than that of the ascospores. The latter have always remained brown, no matter on what medium grown ; the former have not always devel- oped a red color. In a few instances the writer has succeeded in grow- ing jierithecia of the cowpea fungus to maturity, and the shedding of ascospores, without the development of any red color in the peridial wall. In a number of other cases the red color has been very slow to appear. It is thus established beyond doubt that the color of, the peridium of a species may be entirely a matter of the particular sub- stratum on which it hai^pens to be growing. The principal diflerences observed in my cultures are given below : (1) On steamed potato the fungus fruits copiously, the color of the peridium comes quickly (sixth to ninth day), and is first roseate, then bright coral red. The ascospores are mature in two to three weeks. By the sixth day the substratum has changed from white to dark lead gray. In one instance only (out of a great many) on the upper (dry) end of a potato culture 10 days old many perithecia were still nearly colorless. (2) On pearl tapioca steamed with distilled water the fungus grows slowly^ and fruits sparingly; the color of the peridium comes slowly, being first x)ale red and finally black. At the end of 2^- months most of them were destitute of necks, but a few of the black perithecia in one tube had distinct necks. Some of the blackest were removed and examined microscopically. They contained no ascospores or asci. The peridium was brown, but normal in structure. The interior of the perithecia was very full of oil. Some substance necessary to the for- mation of ascospores appeared to be wanting in this substratum. (3) On yellow banana steamed with distilled water the fungus grew slowly and fruited only after several weeks, but then copiously. At the end of 5 days there was not one- fiftieth as much vegetative growth as in corresponding tubes of potato. The perithecia were pale j^ellow- ish red. None were bright red, and some were nearly colorless. These remarks apply to two cultures made i^ovember 13, from the same culture, and examined December IG. In other cultures the perithecia were ochraceous buff at the end of 20 days. On Jannary 23 the two tubes of November 13 were reexamined and found to be quite unlike. In one the perithecia were now coral-red, had decided necks, and were discharg- ing ascospores— 4. e., j)rotruding them through the ostiolum in brown masses. In the other tube, the shape and the color of the perithecia were both aberrant. This culture was covered all over with perithecia, - but these were globose and destitute of necks; a few were pale coral- red, but most were yellowish brown to dark brown. In the darker ones the structure of the i^eridium could not be made oat without crush- ing the perithecium. There were no ascospores in this culture^ not ' The amount of growth at the end of 8 days was not one one-thousandth as much as on potato. 25 even in the dark-brown peritbecia. One of tLe darkest peritbecia was crushed in water. It contained no asci, but every cell was filled with oily looking- globules which blackened with osmic acid. Here we have the substratum causing change in color, change in form, retardation of the production of asci, and the storage of large quantities of reserve material. Most of the old banana cultures resembled this one rather than the preceding. The dark brown or black color was lodged in the walls of the peridium. The oily granules gave a milky appearance to the crushed contents. In some of these tubes certain microconidia were also observed to have changed into round spores. In all, 17 banana cultures were under observation, and for a long time. The one first mentioned seems to have possessed a trace of some substance wanting in the others, and which finally enabled the fungus to fruit normally. (4) On Spanish onion steamed in much distilled water the fungus fruited slowly, but finally quite freely, the peritbecia being pale red to bright coral-red. (5) On commercial (alkaline) cornstarch steamed in distilled water the fungus grew sparingly and fruited sparingly. At first the peritbe- cia were pale red to coral-red, but after a month they were dark — i. e., almost as black as on the tapioca. Even after 2-^ months there were very few peritbecia. Some were i)ale red, but most of them were dark brown, especially the older ones. There were no asci in these dark- brown peritbecia, but great quantities of oil globules. The mycelium was also modified. The latter was extremely variable in thickness (1 to 13 /<), irregular, much inclined to form swollen places and globose ends, or moniliform chains or clumps of big rounded cells. It was also densely packed with refringent globules. In two of the black peritbecia which were crushed in water the peridial wall was brown under the microscoi)e, but normal in structure. (C) On crushed wheat steamed in distilled water the fungus fruited quickly and copiously, but the peritbecia were duller red than on potato. In 8 days the surface was densely packed with dull-red peritbecia. At the end of a month some of them which were discharging ascospores were nearly colorless, while others, also discharging ascospores, had a tinge of purple (not red). Other large clumps, which appeared to be full grown, were entirely without color — i. e., a dirty white. Most, bow- evef, were colored on this substratum. At the end of 2^^ months there was a copious protrusion of ascospores. Nearly all of these were glo- bose, brown, with a thick wrinkled epispore, and were 11 to 13 // in diameter. The interior of these spores was guttulate. A very few of the ripe spores were: (a) smooth, (b) broadly elliptical, (c) ovate. Sev- eral hundred were examined in vain for a septum. At this time the walls of many of the peritbecia, protruding spores, were nearly color- less — very pale red or yellowish white. (7) On hominy steamed in distilled water the fungus fruited slowly 26 aud ill moderate abundance. In 16 days the entire substratum was dark purplisli red ; the perithecia, yellowish red. At the end of a month the perithecia were of good size, but many of the ri])e ones (discharg- ing brown ascospores) were colorless or pale yellow and the rest were yellowish red. i^one were coral-red or bright red of any shade, and the majority were pale yellowish white — i. e., different enough to be considered another species, as species are often made. After 2^ months one of these cultures was still very interesting. The cylinder of hominy, which was now about 3 centimeters high, had shrunk away from the walls of the tube aud its surface bore about a thousand normal-shaped perithecia. A majority of these had jirotruded asco- spores which adhered in irregular brown masses to the top of the neck around and over the ostiolum. The striking fact was that not one of these ripening and rijje perithecia was red; all a- ere pale or dirty yel- lou'ish ichite, including those crowned by the ascospores. Another culture of the same date and kind was exactly like this one, except that it contained about twice as many perithecia, not one of which was red; all were dirty or pale yellowish white. Many were shedding asco- spores, but not so large a number as in the preceding tube. Nothing could be plainer than that the fungus was unable to extract a red color from this substratum; still it was able to ripen its ascospores normally. (8) The color of the peridium was very slow to appear in perithecia on alkaline agars, and on the same with the addition of cane sugar. After 75 to 85 days, only a few were bright red, while many were color- less or only faint roseate or yellowish. (9) On rice steamed with distilled water (5 drops of very dilute hydro- chloric acid added to each test tube) the substratum was dark purplish red at the end of 16 days, except for a few white grains which pro- jected above the rest. The perithecia were numerous and full size, but ranged from colorless to yellowish red. None were vermilion or coral- red. On the surface of a thick rice gruel no perithecia had formed at the end of 24 days. The body of the gruel, which was in a small Erlen- meyer flask, had become dirty white; its surface was mottled brown and pale rose color. At the end of 38 days the body of the rice was light ecru drab, with possibly a trifle of rose color in it. The surface was thickly sj^otted with brown flecks, the largest of which bore numer- ous reddish-brown perithecia. In a few places white hyj)hai were pushing out into the air in small numbers. At the end of 65 days the fungus in this flask had formed a nearly continuous dark-brown sur- face mat, which bore numerous j)erithecia, some of which had shed ascospores. The color of the peridium varied from pale yellowish brown to reddish brown, and even to dark brown. The body of the rice was rather feebly stained — vinaceous cinnamon mixed with a little vinaceous pink. (4) The development of the jyerithecia. — By the use of strongly alkaline media the writer found it possible to entirely prevent the formation of 27 peritliecia in the cowpea fungus, although the vegetative growth and the rorniatiou of the micioconidia continued. This is the more remark- able because on all ordinary media the fungus showed tlie strongest tendency to form perithecia, even when it was unable to ripen its ascospores. The results obtained are as follows: (1) In 92 days, on strongly alkaline beef-broth peptone agar, copi- ously inoculated with fresh ascospores, there was no trace of perithecia. This media consisted of two test tubes of standard nutrient agar (stock 82a). One of them, which contained 10 c. c. of agar, received 10 drops of a saturated solution of sodium carbonate (temperature 25° C), and the other, which contained 8 c. c. of agar, received 4 drops of the same soda solution. Each was resteamed on 3 consecutive days. Boiling with this alkali caused a heavy precipitate in each tube and a dark color in the agar Avhich received the most soda. Growth appeared first in the tube which received the least alkali. Neither showed any for the first 3 days. On the sixth day the slant surface of each was spotted all over by tlie fungus. Growth, however, in each was sparing, considering the number of ascospores used and the time which had elapsed. In the tube which received the least alkali the growth was whiter, and there was more fungus in the air than in the other tube, where the growth was mostly in the substratum. At the end of 10 days, in the tube which received the most alkali, the surface was gray-white, with scattered white flecks. The growth of the fungus was still mostly in the substratum and was not copious enough to hide the surface of the agar. The tube which received one-half as much alkali presented the same general appearance. At the end of 2G days both tubes were almost exactly alike ; the entire surface of each was covered with a gray- ish stroma sparingly flecked with white. In other words, the growth of the fungus, Avhile not so prompt as on neutral or feebly alkaline agar, was not prevented by the strong alkali. At the end of 51 days in both tubes the mycelium showed a marked tendency to break up into iso- diametric, or nearly iso-diametric cells, constricted at the septa and rounded in the middle. In both tubes microconidia were now abun- dant, and it was thought that possibly perithecia might develop later on, as there appeared to be tiny hyph^e-complexes in some of the mycelium, which had crept up out of the alkali bed and was now grow- ing on the walls of the tube. These cultures were started November 30, and up to January 24 they were exactly alike; that is, the surface was covered all over by the fungous stroma, which was grayish white and bore numerous micro- conidia, but no perithecia. On this date 2 c. c. of a sterile 2-per-cent solution of malic acid was pipetted into the tube which had received the least alkali. In 10 days 1,500 unripe hut well- developed and bright coral-red perithecia made their appearance. The whole surface of the corresponding tube (alkaline agar) was covered with a grayish-white sheet of fungus, mostly in the agar (stroma), and it still bore an abun- 28 dauce of microconidia, hut tliere n-as not a shu/le perithecium. Up to the ninety-second day, when the exj)erimeut was lost, not a single perithecium had formed on this alkaline agar. No contrast could be more striking, except possiblj' the following : (2) On the best medium known to the writer for the abundant and rapid production of perithecia, viz, potato steamed in distilled water, the development of perithecia was altogether prevented by the addi- tion of carbonate of soda. A copious stroma was formed, and there was an abundant development of microconidia, hut in 76 days there teas no appearance of perithecia. These cultures were on potato cylin- ders steamed in test tubes with 2 c, c. of water and 1 c. c. of a satu- rated solution of sodium carbonate. A second set of tubes was prepared in the same way, but received less alkali (about one-half as much). For comi^arison, cylinders cut from the same tuber were steamed in a strong malic-acid water. The acidified cylinders were white; those steamed with the alkali were uniformly brown. The three sets were inoculated at the same time (December 16) copiously with ascospores from the same culture. On the fourth day there was a much better growth in the acid tubes; growth in the alkaline tubes was sparing and only to be detected with a hand lens. On the tenth day the fungus in the alkaline tubes had made a pure white but very restricted growth, varying iu area from one-half to IJ qcm. In the tubes which had received the least alkali there was a stroma and an abundance of microconidia, but in none of the four was there any trace of perithecia. On the malic-acid substrata the fungus had now entirely covered and hidden from sight the whole of the large potato cylinder with a wrinkled stroma. This bore in each of the four tubes several thousand perithecia, which were already rose color or bright coral red. The acid appeared to have favored rather than hin- dered their formation. On the seventeenth day the entire surface in each of the acul tubes was so thickly studded with bright coral-red perithecia that there seemed to be no room for any more. These tubes were now no longer acid, although strongly so when inoculated. Each of the four cylinders was tested. There was no trace of acid reaction in two, and in the other two it was extremely faint and limited to the bottom part of the cylinders, which was the last part to be invaded by the fungus. It would therefore appear that the fungus consumed it as food. On this date, on each of the alkaline cylinders, the fungus had formed a stroma, which was gray-white, wrinkled, and mealy from the jiresence of great numbers of microconidia. In the tubes which received the most alkali the stroma covered about one- fourth of the surface and there were no perithecia. In those which received less alkali it covered about four-fifths of the cjdinder in one tube, and only about one-twentieth in the other. In both of the latter perithecia were forming in small numbers, and were first visible on the fifteenth day. The tube containing the smaller stroma was now opened 29 aud the cylinder removed. It suielledlike a soap barrel. The interior of the cylinder and of the stroma were alkaline, and also the beads of water excreted from it. There were very few i)erithecia. On the twenty-fifth day the remaining- tube, which received the least alkali, was fruited rather sparingly, and thepcrithecia were red, but not ripe. In places the stroma was brown and destitute of perithecia; in other places it was grayish-white and destitute. In the tubes which received the most alkali the potato was now covered with a coi)ious, puffed-up, tough, wrinkled, extensive stroma, which had become brown (color of of ascospores) since the last record. Some of this stronui was dug out and examined. The brown color was lodged in the walls of a closely felted mycelium. From this brown mycelium projected many short, colorless hyi)hu', and these bore numerous non-septate, colorless, ellip- tical microconidia. Here and there were small masses of pseudo- parenchyma, probably the beginnings of perithecia, but nothing large enough to be distinguished by shape, even under the compound micro- scope, as perithecia by anj'one not conversant with the method of origin of these bodies. In other words, the alkali did not hinder a copious development of stroma and an abundant formation of microconidia, but did hinder the growth of the perithecia. At the end of 47 days, dur- ing which there had been no development of ])erithecia in either of the tubes which received the larger (piantity of alkali, one of them was opened and 4 c. c. of sterile 2 per cent malic acid water pipetted into it. At this time the potato was covered all over and hidden by a copious wrinkled stroma, which was brown on one side and about normal in color on the other. The next day (February 1) the acid was entirely neutralized, and the fluid around the cylinder still gave a dis- tinctly alkaline reaction, although less than before. The mouth of the tube was now flamed and the fluid poured out. On February 5, as soon as it could be prepared, 5 c. c, of sterile 1 per cent malic acid water was added to this tube, i. e., enough to cover the cylinder. On testing with litmus after adding, the fluid gave a strong acid reaction. On February G there was no reaction to blue litmus, but a slight alkaline one with neutral litmus. An additional 5 c. c. of the 1 per cent malic acid water was therefore added, aud in a few minutes bubbles of gas again began to arise. On February 7 the fluid was again tested, found plainly acid, and poured oft". The cylinder was not, however, fully neutralized, for on February 10 the small amount of fluid in the bottom of the tube was again plainly alkaline. The cylinder was therefore covered once more with G c. c. of sterile 2 per cent malic acid water. This was poured off' after 24 hours aud the tube set away. Up to February 15 there were no perithecia visible in this tube, although it had been 62 days since the tube was inoculated and 5 days since the last of the alkali was removed. Ten days later, however, the stroma (both the brown and the yellow jiarts) was slightly frosted over with white hyphae, recently formed, and bore several hundred pale coral-red peri- 30 thecia, while many others were developing and were still white. The compauiou tube was free from perithecia and remained so until the end of the experiment (76 days). At this time the fungus in the tube which had received only one-half as much alkali had developed perithecia all over the potato, but they were still immature and either colorless, yel- lowish, or pale brown. (3) In 40 days on rice with bicarbonate of soda there was no pro- duction of perithecia, although the tube had been copiously inoculated with ascospores. The tube contained 10 c. c. of rather dry boiled rice, to which was added 2 c. c. of a cold saturated solution of the sodium bicarbonate in water. In 5 days there was only a very feeble white growth. At the end of 40 days the fungus had grown sparingly over most of the rice grains in the upper half of the tube, and micro- conidia were present, but the rice had not changed color and no peri- thecia could be found. OTHER BIOLOGICAL PECULIARITIES. In addition to what has already been said, there are a few other peculiarities which may here be mentioned. The fungus in all its varieties is a strict aerobe. Up to this time the writer has never been able to find any substance in the presence of which it will grow in the closed end of fermentation tubes. It seems entirely unable to obtain its respiratory oxygen from food substances. The melon fungus is able to obtain its nitrogen from asparagin. In carbohydrate foods nearly destitute of nitrogen, and in which the fun- gus could make but a very feeble growth, it at once made an excellent growth on adding a small fragment of asparagin. Old rice cultures of the melon fungus gave off" a peculiar, rather pleasant, aromatic odor when boiled in water. On boiled seeds of the cowpea the melon fungus grew vigorously and developed a peculiar musky, pungent odor, which was observed as soon as the tubes were unplugged and was quite unlike that just mentioned. The melon fungus grew vigorously on potato and nutrient agar which was sterilized with a large quantity of sulphur dust. The melon fungus grew on nutrient agar in the presence of large quantities of caustic lime and of carbonate of lime. The melon fungus also grew (at first very slowly and in small patches) in a tiask of boiled rice in the presence of very large quanti- ties of sodium carbonate (one-half pound of rice cooked with about 300 c. c. of distilled water and 100 c. c. of a saturated solution of sodium carbonate, saturation temperature 25© C). The quantity of alkali was sufficient to make the rice quite yellow. In a test-tube culture, the cowpea fungus grew on a cylinder of potato which had been boiled with 2 c. c. distilled water and 1 c. c. saturated solution of sodium carbonate (25° 0.). After steaming, one-half of the fluid was poured off, so as to expose part of the potato to the air, and the surface was inoculated with ascospores. 31 Both the melon and the cowpea fungus grew in nutrient gelatin (_|_40, +20, 0, and — 20 of Fuller's scale), but without liquefaction. The melon fungus grew rapidly on glycerin agar. In Dunham's solution with 2 pir cent cane sugar and 0.2 per cent sodium carbonate the melon liingti.s made more growth than the cow- l)ea fungus (ascospore strain). After 13 days the former filled the tubes (10 c. c. portions), while the latter had made only a trifling growth on the bottom of the tubes. In each the fungus was browned. Both the melon and cowpea fungus made a very insignificant growth on dried ligs steamed in distilled water, the very sweet substratum ajipearing to be unsuitable for growth. In acid media, such as the Juice of ripe Concord grapes, the growth of the melon fungus was much retarded, but it finally overcame the inhibiting substances. The melon fungus grew well in sterilized horse dung and feebly on rotten wood (soft maple destroyed by wood fungi). Bordeaux mixture sprayed upon young melon vines in no way checked the spread of the disease. Carbonate of copper mixed with carbonate of lime and put into the hills at planting time, or soon after, did not protect either cotton or melon plants. HOST PLANTS. Neocosmosjjora occurs on cotton {Gossypiuni herhaceum and G. Barba- dense), watermelon {Citrullus vulyariH), and on cowpea (Fiy May 25 melons enough had wilted to make a total of about 100 cases. The cotton and cowpeas were free from wilt and growing very satisfactorily, considering how closely they were planted. By June 1 there had been 166 cases of melon wilt on this bed. All of these plants were examined microscopically, and in every one there was an abundance of the Fusa- rium in the vessels of the tai)root or stem, or both. In the recently wilted plants the iiingus was restricted to the vessels; in those which had been wilted some days, but were not yet dry- shriveled, it was also in the parenchyma, but had not reached the surface. Only 16 healthy melons remained, and these subsequently contracted the disease and died. In digging and pulling out the diseased melons for examination the roots of the cotton and cowpea plants must have been considerably broken and disturbed, but neither in this way nor by the aid of the Thielavia or of another fungus destitute of fruit, but resembling a Pythium, was the melon fungus able to find its way into the cotton or cowijea stems. Up to August 8, when the experiment was discontinued, all of the cotton and cowpea plants remained free from this disease. At this time they were large plants, and the cotton was suffering considerably from crowding. ONE FUNGUS, OR THREE? It is my intention to repeat and extend the ascospore inoculation experiments as soon as time permits, and also to settle more definitely by means of additional cross inoculations whether under any circum- stances the fungus from any one of these plants will ever transmit dis- ease to the others. Much time has already been devoted to this problem, which is one of great practical imi)ortance. From certain cultural peculiarities of the fungi, from the uniformly negative results of the hothouse experiments and from certain field observations, it now looks as though these were separate diseases due to closely related but not identical organisms. This is also the opinion of some very well- informed growers. One of these field observations may here be given. In July, 1894, the writer examined a large field of upland cotton belonging to Mr. T. S. 40 Williams, of Monetta, S. C, without finding any cases of the cotton wilt, and Mr. Williams stated that there had never been any in it. This cotton was planted on land where watermelons had wilted badly in 1893. It was therefore on a large scale an experiment similar to those described above. In 1895 the field was again planted to cotton, alter- nating with cowpeas — i. e., 3 rows of cotton and 3 rows of cowpeas. In September of that year the writer again examined the field quite carefully, and with t*he same negative result. No cases of cotton wilt were to be found, but there were hundreds of wilting and dead cow- pea vines, and the latter bore the external conidia beds of the luuulate- spored Fnsarium quite regularly, and these were arranged in the manner shown in PI. I, 11, 12. The melon fungus was undoubtedly present im the soil, but there were no melon vines on the field by means of which to establish this fact conclusively. At my request Mr. Orton reexamined this district in the summer of 1899. He found an abun- dance of melon and cowpea wilt, but none of the cotton disease. Morphologically, on the contrary, so far as I have' been able to determine from the careful examination of a great many specimens, this fungus is specifically identical on all of these plants, the only doubt being whether it may not have varied enough physiologically not to be transmissible from one host to the other. There are slight variations in the length of the beak and in the size and color of the perithecia on the different host plants, but these are not constant. The ascospore variations are also not constant. As already stated, the ascospores of the form on watermelon are usually smaller and more decidedly elliptical, and a larger proportion are smooth (perhaps because small). When this was discovered an earnest attempt was made to distinguish two or more species, but further studies developed the existence of all sorts of intermediate forms and sizes and left no morphological standing ground for any such conception. For example, perithecia were found on the watermelon at Monetta, S. C, in Septem- ber, 1895, which contained globose wrinkled ascospores 12// in diam- eter; ascospores 12 by 13// and 12 by 14 /< were also observed. Other perithecia on the same root bore elliptical smooth spores. Indeed, ascospores from the same perithecium may be globose or elliptical, wrinkled or smooth. Ked perithecia taken from the roots of wilted, dead cowpeas, at Monetta, September 10, 1895, were indistinguishable in color, size, and shape from those found on watermelon roots, and bore ascospores of the following sizes: 9 by 10, 9 by 14, 10 by 10, 10 by 12 (wrinkled), ]0 by 13, 10 by 14 (smooth), 12 by 12 (wrinkled). Ascospores from perithecia which grew on the roots of sea-island cot- ton on James Island were 10 by 10 (smooth and not many so small), 10 by 12 (smooth and wrinkled), 12 by 12 (wrinkled), 13 by 14 (wrinkled). Owing to variability, it is likewise impossible to distiugnish the fun- gus from the various hosts either by means of the mycelium or by the external or internal conidia. At most the differences can be scarcely 41 more than varietal — i. c, such as miglit be induced by the long-con- tinued use of different substrata. This is true even if the fungus can not be transmitted from one liost to another, and a suflicieut number of experiments have not yet been made to enable one to declare with- out reserve that this never takes [)lace under any circumstances — e. g., with help of some other fungus, or under peculiar conditions of environ- ment not yet discovered. At present, therefore, nothing remains but to consider the fungus as one species and to record the forms on the other host plants as varieties. Possibly it may finally turn out that they do not deserve even this rank, but the writer does not now feel justified in giving tliem any less. VITALITY. From its vitality under adverse conditions, its ability to live in the dung heap and in the soil, and the ease with which it may be cultivated on all sorts of artificial media in the laboratory, this fungus must be regarded as a serious eneuiy to agriculture. While we know it only on the i)lants mentioned, it is probably capable of attacking other species, and ought certainly to be expected and looked for on other hosts.^ The length of time the fungus will remain alive in the earth is remarkable, and adds greatly to the difficulty of combating it. Why it should ever disappear, any more than a bad weed, when once estab- lished in a cultivated soil, is not dear. It should certainly be regarded as a weed, and one the eradication of which presents unusual difficulties. In extensive field culture it has been found unsafe to plant lands which have once suffered from it until after a lapse of several years— five to seven, according to south Georgia melon growers, and certainly more than three, as shown by a 7-acre field test in South Carolina, which came under my own observation in 1895. Only 3 wagonloads of melons were obtained from the whole field. The melon fungus has lived a year in the dried-out soil of pots used in my greenhouse experiments, and a very similar fungus parasitic on cabbage remained alive in dry soil three and one-half years.^ The 1 Since this was written Mr. Orton has found the fnngns on James Island, South Carolina, in a weed, the Cassia ohtusifolia L. Not many plants were attacked, hut the external symiitoms were identical with those on the cotton and cowpea, while the walls of the vessels of stem and root were stained hrowu and the lumeu was filled frequently with mycelium, abstricting the typical microconidia, and sometimes browned, as in cotton. -The writer has just concluded an experiment with a similar looking and acting cabbage Fusarium. In parts of New York, Virginia, and Maryland this fungus has troubled the market gardeuers, in some cases renderiug impossible the profitable culture of cabbages on large and fertile fields — e. g., a field in New York. which for- merly yielded from 90,000 to 95,000 heads of marketable cabbage for each 100,000 plants set out, can now l>e depended upon for only about 30,000 heads; on a lield in Maryland which formerly yielded good croi)s, the cabbages were so badly affected this year that the ground was replowed and planted to other crops in the middle of the season. The symptoms in the cabbagi- are slow growth, refusal of the heads to form, a sickly color, aud the premature shedding of the lower leaves, from the axils 42 melon fuugus lived in a dried-out condition in one of my agar tube cultures tbrec and one-half years; in another it was alive at the end of ten months and twenty-three days; in a third trial it was found alive in 7 out of 8 test-tube cultures which had been in a warm, dry place in a dried-out condition for nearly two years. These last tubes were inoculated at various dates between July 26 and October 8, 1894, and the test was made January 19, 1897. The ascospores of the cowpea fungus remained alive iu a dry condition 16 mouths.^ SYMPTOMS PRODUCED, ETC. The gross symptoms in the watermelon are sufficiently shown on the accompanying plates (Pis. VII, VIII, X). They are those of a plant of which vshort sprouts frequently push out. This disease was first studied hy the writer in 1895. It a+tacks the plant in the same way as the melon fungus — i. e., from the soil — and destroys it by plugging the water ducts. It first produces iu the ves- sels of the living plant great numbers of microconidia (8 to 13 by 2 to 4 //), and then macroconidia on the surface, in the same way as the melon fungus. In July, 1895, a quantity of soil was obtained from one of these badly infested fields (near Albany, N. Y. ) and was stored away in a dry basement for three and one-half years. It was then removed from its original packings, put iuto a clean pine box made from freshly planed lumber, planted with three varieties of cabbage, placed in an upper window of an unoccupied laboratory room, and watered with distilled water. Some of the plants were attacked by Pythium and others by this Fusarium, which was found fruiting in the vessels. The checks did not contract the disease. The conditions under which the experiment was made point unmistakably to the soil as the source of the infec- tion in case of these plants. This experiment, in connection with those which have been made on the watermelon, renders it probable that all parasitic soil Fusaria are alike in being very resistant to tlrought and other conditions unfavorable to vegeta- tive life. Embolisms of the vascular system due to fungi of the form-genus Fusarium are now known to the writer in plants of many different families. (See partial list on p. 43.) Most if not all of these fungi enter the plant from the earth. The fungi occurs so regularly in connection with these diseases that we are warranted in assuming them to be parasitic, although as yet in most cases no infection experi- ments have been instituted or brought to a successful conclusion. Judging from the results obtained with the watermelon and cabbage, it appears extremely prob- able that iu the Cephalosporium and Fusarium stages of a variety of Nectriaceous fungi we have to deal with a large group of destructive soil parasites the very existence of which, in the earth, as active jjarasites, was not suspected until very recently — i. e., until the announcement of my results in 1894. (Am. Asso. Adv. Sci.) ' Some additional tests were begun September 28, 1899, as this bulletin was passing through the press. The watermelon fungus was found to be dead iu each one of the 10 bread cultures already mentioned (p. 16) and in a few other cultures of the same age (hominy, cowpeas), but was still alive in numerous test-tube cultures of horse dung. These were made October 27, 1894 (from internal and external conidia), and have been dried out for at least four and one-half years. Twenty-two tubes were tested, and the fungus was found alive in each one. Four test-tube cultures of the cotton fungus (from internal conidia) were also tried at the same time, and the fungus was found to be alive in each one. These 4 cultures were 5 years old. They were made on sterilized stems of the watermelon, October 3, 1894, and have been dried out and in a dry laboratory for at least four and one-half years. The cowpea fungus (ascospore strain) was dead iu each one of the 12 banana cultures made December 6, 1895. 43 transpiring freely and insufficiently supplied with water, althouyli at the same time there is an abundance of moisture in the soil. The uniformity with which this fungus first seeks out the vessels of the plant is very striking (PI. 1, 10, PI. IV, and PI. V, (i). This explains the sudden wilt of tlie foliage. The water ducts are clogged to such an extent that they can not function. That this wilt is attributable to lack of transpiration water, brought about by partial or complete clog- ging of the vessels, is shown by the fact that large plants which have begun to wilt frequently recover for a day or two if a rain sets in and the air continues moist. During such weather the progress of the dis- ease in old vines is almost at a standstill, but it recommences when the sun shines out and the moisture of the air is dissipated. The mechan- ical nature of this obstruction is also shown by the fact that collapsed leaves frequently recover their turgor on cutting stems above the fun- gous plugs and plunging them into water, e. g., on cutting across the second internode when the plugs are confined to the hypocotyl and tap- root. It is also shown by the tiict that the terminal portion of shoots which have collapsed and the basal vessels of which are plugged by the fungus (as shown by the death of all the lower leaves and of all the other branches) may survive for several days, if the weather is not too hot and dry, provided they are attached to a large melon from which they can draw a certain amount of water (PI. VII, -5, right-hand branch). Finally, it is much easier for the fungus to plug all of the vessels of a small plant than of a large one, since in the taproot it has to make but a little growth through tender tissues to accomplish this. This ex- plains why yoiiufi vines freiiuently wilt even during rains or when the soil and air are very moist. The leaves of the cowpea usually unjoint and fall off, leaving the green stems bare. Often some of them become yellow and fall without first showing signs of wilt, as Professor Atkin- son has recorded in case of the cotton disease. The watermelon leaves do not become yellow or detach, but wilt suddenly in large numbers and shrivel, so that a large, healthy-looking vine may lose all of its foliage in twenty-four to forty-eight hours (contrast PI. VI and PI. VII). The cotton plant is less susceptible, contains, as a rule, less of the fun- gus, and often recovers partially, so as to produce some fruit. In such instances the fruit-bearing stalks push out of the base of the stem and finally hide more or less completely the main shoot which has been killed by the wilt. The xylem of the diseased plants always becomes brown (PI. 1, 10 a, b), and in case of the translucent stem of the cowpea this stain shows through the green bark, giving an unusually dark appearance to the still living stem. This browning of the xylem appears to be common to all plants attacked by parasitic soil Fusaria, e. g., cotton, okra, cow- pea, beans, watermelon, squash, potato, tomato, eggplant, red pepper, sweet potato, cabbage, carnation, asparagus, pineapple, and others. This browning begins in the walls of the larger vessels, and often it is confined quite exclusively to the xylem for a long time, the pith, bark, 44 and pliloem remaining free from stain. In tbe cotton fungns the older mycelium itself, inside the vessels, is frequently stained yellow or brown. Browning of the mycelium of the melon fungus inside the vessels of the plant has not been observed, but it occurs in pure cultures on boiled melon stems, etc. Once started, in cotton at least, this browning may extend long distances through the woody parts of the stem with very little fungus to help it on. OTHER WILT DISEASES. This disease should not be confused with the cotton-root rot of Texas, from which it appears to be distinct,' nor with another wilt of cotton, cowjjeas, etc., common in parts of Florida and first described by Prof. P. H. Eolfs. This is associated with a fungus which attacks many kinds of plants, wild and cultivated, enveloping the base of the stem externally with a copious, white, rather straight mycelium, bearing on its surface large numbers of small sclerotia which are first white, then fulvous, and finally dark brown and smooth. When grown on nutrient media (agar), the fungus reproduces itself by another crop of sclerotia, and so on indefinitely (a year in my laboratory). These sclerotia are mostly 0.8 to 0.9 by 1.2 to 1.3 mm. RELATIONSHIPS. The perithecium of Neocosmospora much resembles a medium- sized bright red Nectria, and if gathered in an immature condition would nat- urally be placed under ISTectriella, as the spores are smooth and colorless and there is no indication of any septum, even in the most elongated or in very old spores. When ripe, however, the ascospores are distinctly brown, and the fungus clearly belongs to a new genus. It possesses some of the characters of both Nectriella and Melanospora, but is dis- tinct from either. On the whole it seems to be nearer to Nectriella than to Melanospora, although the spores are brown. It is not very closely related to Melanospora as originally established by Corda and defined in Saccardo's Sylloge Fungorum. It is widely different from MelanoHpora chionea, M. leucotricJia, M. zamia', M. vitrla, M. zobelii, M. coprina, M. laf/enaria, M. 2mrasitica,, and all other Melanosporas and Sph.erodermas which the wiiter has been able to examine or to find figured. The ascos^^ores of this fungus are particularly unlike those of Melanospora. In the latter they are smooth, often lemon-shaped, or even apiculate, and frecjuently obh(iue or fiat on one or both sides. The general appearance of Melanospora indicates that it might perhaps be properly excluded altogether from the Nectriaceous fungi, its affinities 1 Four hundred cotton plants killed by this disease were received from two differ- ent localities in Texas in 1895. Pamniel's Ozoniuni was present, bat not a trace of any stage of this fungus could be found. These plants came from regions much subject to root rot. 45 beiu;^ witli tbe Sordariaceti', as Scbni'tt^r pointed out in liis Kryptoga- inen-Flora von Sclilesiens. The development in the cowpea fnnous of peiithecia with and without beaks on mycebuni derived from a single asc'ospore makes it probable that the same thing oceurs in other genera and tends to «',outirm Winter's view of the untenability of the genus Sphieroderma, wliieh was erected to include those forms of Melanospora destitute of a beak. The development of the perithecia of this fungus, either in the soil near the host or on the surface of the latter (the more common way), or under the substratum in rifts or cavities of the host, shows that this character has no particular value and leads one to sus- pect that the genus Ilyponectria may also have no sound physiological or mor[)hological basis. This fungus is most nearly related to the Cosmospora of Raben- horst, which is a good genus and should be reestablished. It differs from Cosmospora chietly in having non-septate ascospores and a wrinkled exospore,the former having Iseptate ascospores with a papil- liite or verrucose exospore. ^ly cultures have shown beyond doubt tliat the ascospore is less readily modified by the substratum than any otlier part of the fungus, and therefore the most important for purposes of classification.^ 2 These two genera may 1>6 defined brielly as follows: Cusmoftwra Rabh. (emend.). Peritbecia as in Nectrla (red in the known species). Asci numerous. Ascospores 8 in one row, brown, oblong elliptical, 1-septate, usually more or less constricted at tbe septum, with a distinct papillate or verrucose epispore. Paraphyses present, inconspicuous, broad, loosely jointed, unbranched. Couidial stages unknown. 1. C. coccinea Rabh. 2. C. Cameroensis (Rehui). Syn. Sphieroderma Cameroensis Rehm. Neocosmospora. Perithecia as in Nectria (bright red in the known species). Asci numerous. Asco- spores 8 in one row, brown, globose or short elliptical, continuous, with a dis- tinct wrinkled epispore (the latter sometimes wanting in the smaller spores). Paraphyses present, inconspicuous, broad, loosely Jointed, unbranched, con- sisting of about 5 cells. Three conidial stages, viz, Cephalosporium, Fusa- rium, and Oidium. 1. N. vasinfecta (Atk.). On cotton (parasitic). Perithecia as below. Spores usually globose, wrinkled, generally about 10 by 10 //. a. Var. tracheiphila (Smith). On cowpea fparasitic). Perithecia quite variable, but mostly 250 to 3.50 /.i tall by 200 to 300 /< broad, with or without a short neck; on the dead roots or in tbe soil over them Spores usually globose, wrinkled, mostly 12 by 12 //. /3. Var. nivea (Smith). On watermelon. Very actively parasitic. Enters the plant from the earth and plugs the vascular system, causing a sud- den extensive wilt of the foliage. Perithecia as above. Spores globose or elliptical, wrinkled or smooth, generally smaller than in the preceding and more often elliptical, but variable. 46 SCIENTIFIC NAME. For the present, at least, this fungus may be designated as follows: Neocosmospora vasinfecta (Atk.) Syn. Fusarium vasinfectum Atk. On cottou. Probably also on okra. Parasitism not proved. Genetic connection of various spore forms not proved. Cblamydospores not observed. a. Var. tracheiphila. Syn. Nectriella traclielpMla, Erw. Sm. On cowpca. Parasitism not pioved. Genetic relationshiji of ascospores, macro- conidia and microconidia established. Cblamydospores not observed. /?. Var. nivea. Syn. Fusarium niveuvi, Erw. Sm. On vratermelon. Not known to occur on other Cncurbitacetc' Parasitism fully established. One of the most destructive soil parasites known. Genetic connec- tion of conidial stages fully established. Peritlieeia identical with the preceding have been found on stems killed by the internal fungus, but the genetic relation- ship has not been proved by exact culture experiuients. By "not proved" the writer does not mean that he has himself any doubt whatever as to the parasitic nature of the cotton or cowpea fun- gus or as to the genetic relationship of tlie various spore forms on cotton and of the perithecia to the conidial stages on the watermelon, but that these points have not been definitely settled by satisfactory infection experiments and by deriving one spore form from the other in pure cultures. Without such experiments the proof remains incom- plete. The field evidence, however, of the parasitic nature of the cot- ton and cowpea fungus is of the most convincing sort, i. e., the fungus is always present in the vessels of the diseased plants, nothing else is always present, and the disease occurs year after year on the same soils. This constant association of the fungus with the cotton and cowpea disease makes it reasonably certain (in the light of what has been accomplished with the watermelon fungus) that abundant infec- tions will be secured at no very distant day. The failures thus far have probably resulted from a greater resistance on the part of cotton and cowpea, or from the fact that the natural method of infection has not been discovered. Possibly the cotton and cowpea may be subject to infection only during germination or early stages of growth, or only ' A destructive disease of cucumbers and muskmelons associated with a Fusarium in the stem bundles has been reported from Ohio by Prof. A. D. Selby. The occur- rence of a similar disease in muskmelons has just been reported from Connecticut by Prof. William C. Sturgis. A Fusarial disease of squashes is known to the writer. Possibly these diseases are due to the watermelon fungus, but the writer has not observed diseases of other Cucurbitaceous plants in regions subject to the water- melon wilt, and until cross-inoculation experiments have shown their identity it is proper to consider these diseases as distinct. 47 when injured in sonic particular way. If these plants are infected chiefly during the seedling stage, then the growth of young plants in uninfected seed beds and their transplantation to infected soils only when they have passed out of the receptive stage would atl'ord relief. This is offered, however, only as a working hypotliesis. All we yet know is that frerpiently many plants of cotton and cowpea on infected land come to a healthy maturity, i. e., there is not such a sweeping general infection as in case of the watermelon. Possibly, however, this is a statement depending on an insufticient number of observa- tions. It is true so far as my observations have extended.^ Melon plants grown for several weeks in good earth, i. e., plants with one true leaf and the second one beginning to develop, are still freely subject to soil infection. The shortest period of incubation observed by the writer has been about C days, i. e., the cotyledons wilted, as shown in PI. VIII, the first or second day after the plants emerged from the ground. The longest period of incubation, or rather lapse of time between infection of the soil and appearance of the disease, was 81 days. In this case the seed was planted and the soil of the pot infected April 25 and the vine showed no symptoms of disease until July 15, when it suddenly wilted. Thirty-one other vines of the same series contracted the disease between June 1 and July 11. The soil of these pots was reinfected May 20, no cases having appeared. In both instances the fungus used was derived from a big, arcuate, several- septate, external conidium. Vine 32, which wilted at the end of 81 days, was 4 feet long. There was an abundance of fungus in the ves- sels and it bore only the small, colorless, elliptical microconidia; these were non-septate, straight or slightly curved, and measured G.5 to 24 by 2 to 4 //. In field culture a period of 80 days frequently intervenes between planting and the first appearance of disease in a given vine. The question of parasitism of the cotton fungus was left an open one by Professor Atkinson, as shown by the following citations: "The Fusarium was considered not to be a sufficiently aggressive parasite to be able to make its way into the ducts of the circulatory system unaided." It is suggested that a dampiug-oflf fungus may first open the way. The only two infections obtained by him with the cot- ton Fusarium were on plants with open wounds made by the "sore- shin" fungus. "The discoloration and disease of the ducts is started I Mr. Orton obserred in 1899 in various parts of Soutli Carolina what had previ- ously escaped my attention, viz, that every cotton plant in the vicinity of dead and dying plants is dwarfed and does not bear as much fruit as it should, even when it shows no symptoms of the disease. This dwarfing is most conspicuous toward the end of the growing season and is associated with the presence of the fungus in the vascular system of the side roots of the plant, the vessels of the tap- root and of the stem being free or very nearly free from the fungus. This may help to explain the statement often made by cotton growers that one stalk in a hill will blight and another will not. In reality, it may be only a question of degree, all being more or less affected. 48 by the injury from the 'damping-off' fimgus." "While I do not wish to be understood as making any positive assertion in favor of the Fusarium being the cause instead of bacteria [which were sometimes found associated with it], I do think the evidence thus far in hand gives greater support to the former view. The Fusarium is invariably found both in cotton and in okra afflicted with the disease. Bacteria are not always found in the diseased tissues, etc." Some experiments by Professor Atkinson also led him to believe the external Fusarium on cotton distinct from the internal one. RESTRICTION OF THE MELON WILT. While no cure is known for this disease, our knowledge of its cause and manner of spread is now sufficiently exact and complete so that certain rules of practice may be given. By carefully following these the farmer will frequently avoid very serious losses. (1) Fields already infested with this fungus must not be jilanted to melons for a long series of years. So far as yet known, canteloupes, cotton, peanuts, cowpeas, soy beans, or velvet beans (Mucuna sp.) may be i)lanted on such fields without danger. (2) Fields free from this disease may become infected by the wash from lands already infested, and probably, also, by means of the dirt adhering to agricultural implements and to the feet of horses and cattle. For this reason, if cattle are pastured on such fields, they should not be allowed to roam freely over uninfected j)arts of the farm, and tools used on such lands should at least be scoured bright before using on other fields. Where uplands are infected the wash from sud- den heavy rain storms should be turned aside, as far as possible, from uninfected lowlands. (3) Inasmuch as the vitality of the fungus is great and the wilting melon vines are full of it, the danger for a half year or more from such vines is very great. All such plants are magazines of infection. They should be pulled while green, stacked with brush, and burned. Large growers of melons could well afford to keep one man in the field all the time for this purpose. The plants should be removed as soon as they show distinct symptoms of the wilt, because at this time the fungus is still confined to the interior of the stems and not likely to be scattered about by the removal, as would be the case a few weeks later when the vines are dry and the fungus has fruited abundantly on their surface. This precaution should not be neglected simply because fields show only here and there a wilted vine, since in course of a few years such fields have been known to become so thoroughly occupied by the fungus as to altogether prevent melon growing. (4) Occasionally the fungus is introduced into the barnyard, so that the dung pile becomes a source of general infection to fields previously free from the disease. This is apt to be the case where "melon hay" is 49 fed or used for bedding.' The writer discovered one striking outbreak of this disease iu South Carolina in 1894 which could be accounted for in no other way, the divsease making almost a clean sweep on the live acres which received most of the manure. If there is the least reason to suspect the manure pile, commercial fertilizers should be used instead. (5) Farmers whose lands have become generally infected are advised to grow other crops on their own fields, and to rent uninfected land from their neighbors for the purpose of melon growing.^ UNFINISHED WORK. It is a source of regret that this bulletin could not be made more complete, but during the period covered by this investigation many other important lines of work have also demanded attention. In the scant time at the writer's disposal for this special work he Las there- fore done as juuch and as well as he could, and must so leave it. So far as relates to the life history of the fungns, the most important things remaining to be done are as follows: (1) Determine time and manner of infection of cotton and cowpea plants. (2) On cotton establish the genetic relationship of the various spore forms, perithecia and external and internal coiiidia. (3) On watermelon obtain experimental proof that the perithecia on the dead stems bear the same relation to the internal and external conidia as they do in case of the cowpea fungus. ' (4) Obtain infections with ascospores on cotton, melon, and cowpea. (5) Connect the okra fungus with the cotton fungus by experimental studies. (6) Determine why it is impossible to grow perithecia from some conidia and easy to grow them from others. Does remoteness of origin from the ascospore interfere? (7) Determine by additional cross-inoculations and field observations whether under any circumstances the fungus from one host plant can transmit the disease to another host plant. (8) Determine how and where the melon fungus gains an entrance into the plant. (9) The ease with which the perithecia may be grown would seem to ofi'er a good opportunity for studying the question of the sexual or non-sexual origin of the ascospore fructification in this group. * Melon hay consists of hay made on inolou fields late in the season, after the melon crop has been harvested. It is composed of wild grasses iDterspersed with dead melon stems. If the latter are inft'sted with this funi^ns, then the dung heap becomes inoculated, and subsequently any laud on whicli this manure is used. ^This advice was given by the writer to Mr, T. S. Williams, of Monetta, S. C, in 1894, with the happiest results, as indicated by the following excerpt from an unso- licited letter written by him iu 1897: "Financially your researches here have been worth thousands of dollars to myself and others. By the information you gave us we were able in a great measure to avoid land likely to die." 4133— i^o. 17 4 50 Many practical questions are left for further investigation, especially all that relates to cotton. EXSICCATI. Specimens of this fungus will be distributed in Ellis and Everhart's Fungi Columbiana (Century 15). These sets include only the fungus as it occurs on cowpea and on culture media derived from this source. These specimens will consist of the following material : [a) Perithecia on stems and roots of the cowpea {Vigna sinensis)', {h) pure cultures of immature perithecia grown in the laboratory from ascospores sown on steamed sterile potato; (c) ripe perithecia from pure cultures on steamed potato; {d) stems of cowpea with the vascular system plugged by the white fungus, which bears microcouidia (these stems were green when gathered and bore no external conidia beds, but in many instances, to my chagrin, the fungus pushed through and fruited on the surface while the stems were drying); (e) external salmon- colored conidia beds (macroconidia) on stems previously killed by the internal fungus and which were dry when gathered. Type specimens have been deposited in the cryptogamic herbarium of the Division of Vegetable Physiology and Pathology, United States Department of Agriculture, and these specimens, which have been selected with equal care and are of equal value, may be regarded as co-types. PREVIOUS LITERATURE. (1) Atkinson. Fusarium vasinfecium. ' Some Diseases of Cotton. Bull. No. 41, Dec, 1892, Agricultural Experiment Station, Auburn, Alabama, pp. 19 to 29, with 3 figures : a diseased leaf, internal mycelium in vessels of cotton plant, mycelium and conidia from cultures. (2) Atkinson. Diseases of Cotton iu "The Cotton Plant," a Bulletin (No. 33) issued by the Office of Experiment Stations, Dep. of Agric, Washington, D. C, 1896, pp. 287-292. Nothing new added. (3) Smith. Fasarium tiiveum. The Watermelon Disease of the South. Proc. Am. Asso. Adv. Science, 1894, p. 289. (4) Smith. NectrieUa tracheiphila. The Watermelon Wilt and other Wilt Diseases due to Fusarium. Proc. Am. Asso. Adv. Science, 189.5, p. 190. (5) Smith. The iiath of the water current in cucumber plants: (5) The result of parasitic plugging of the vessels. American Naturalist, 1896, p. 561. (6) Smith. The Spread of Plant Diseases: A consideration of some of the ways in which parasitic organisms are disseminated. Tr. Mass. Hort. Soc, 1897. APPENDIX. A fungus or supposed fungus, described very imperfectly by L, v. Schweinitz iu bis Synopsis fungoruni Carolinae superioris, and by Fries in bis Systema Mycologicam, as S2)lt(vna gossypii, bas caused tbe writer some trouble and sbould be mentioned bore. It was found on unripe cotton bolls and Fries stated tbat be bad seen dried specimens. From tbe description in bis Systema, making allowance for inexact observa- tions and unwarranted inferences, it seemed at first not unlikely tbat tbe peritbecia found by me on tbe cotton plant migbt be tbe old t^phcc- ria gossypii of de Schweinitz. This view I have now abandoned as untenable. Tbe origin and history of tbis name, wbich must be regarded as a nomen excludendum, are as follows: (1) De Schweinitz. Synopsis Fungorum Carolinae superioris, p. 46. 207. [Spliieria] Gossypii Sz. S. simplex sparsa iuimersa globosa purpureo-incarnata submollis, ostiolo ad super- ficiem elongate gelatiuam fundeute. Vix capsula Gossypii iiivenitur siue hac Sphaeria. Minuta. Nascitur in capsulis Gossypii immaturis, primum profimde immersa, sed ostiolo etiam turn ad superficiera elongato, concolori, unde spargitur gelatina luduresceus, qnse capsulis purpareo- rutroiuquinat. SptiauuliL- non observantur nisi capsula percissa; demum assurguut et superficiem turn exsiccatam granulosam redduut. Tbis species de Schweinitz placed in bis Family VIII, Simplices, Sec. C, Brachystomie, along witb S. putaminum on peacb pits and some others. (2) Von Schweinitz. Syu. N". Am. Fungi, p. 217. Mere mention as follows : *1652. 507. -S^. (jossypn, L. v. S., Syn. Car. 207, F.212, non in Pennsylv. (3) Fries. Systema Mycologicum, Vol. II, 1823, p. 488. 412. S. Gossijpii. Sparsa, submollis, peritbeciis immersis globosis, purpureo- incarnatls, ostiolo ad superficiem elongato gelatinam fundente. Schwein. (!) 1. c. n. 207. Minuta, dubijB affinitatis. Nascitur in capsulis Gossypii immaturis, primum pro- funde immersa, sed ostiolo etiam turn ad superficiem elongato concolori, unde spargitur gelatina indurescens, quj© capsulas purpureo-inquinat. Peritbecia non observantur nisi capsula perscissa, demum assurgunt et superficiem tum exsiccatam granulosam redduut. In capsulis Gossypii copiose. (v. s.) (4) Saccardo. Sylloge Fungorum, Vol. II, p. 457. Saccardo not knowing wbat to do witb tbis species refers it doubtfully to Hyponec- 51 52 tria as E. gossijpii, quoting with a slight rearrangement of words the description of Fries. (5) Ellis. North American Pyrenomycetes, p. 71. Ellis nses Sac- cardo's name, gives an incomplete translation of the Latin description, and adds: We have seen no specimens of this species, hut have received from Prof. F. L. Scribner a Fiisarium on capsules of cotton from South Carolina, which may he the conidial stage. (6) Ellis. Notes on Some Specimens of Pyrenomycetes in the Schweinitz Herbarium of the Academy. Reprint from Proceedings of the Academy of Natural Sciences of Philadelphia, February 21, 1895, p. 11. Sphseria Gossypii, Schw. Syn. Car. 207. This is an obscure thing. The inner membrane of the cotton boll is wrinkled or roughened in drying so as to give the appearance of minute perithecia, but there is no fruit nor even any real perithecia. (7) Curtis. In Dr. Farlow's herbarium, at Cambridge, Mass., is a fragment of cotton capsule labeled "Sphoeria gossypii Schw." This came from the herbarium of M. A. Curtis, who received it from de Schweinitz. On the pocket in the handwriting of Mr. Curtis is a pen- ciled memorandum to the eft'ect that this is not a fungus. Neither the writer nor Dr. Farlow, who examined the specimen with him, could find any Sphiieriaceous or Nectriaceous fungus or any Fusarium spores on this specimen. (8) Dr. Karl Starback, who kindly examined for me the Schwei- nitzian material of this species sent to Dr. Fries, and now preserved in the Fries collection at Upsala, writes that no fungus whatever is pres- ent and adds: " Species Schweinitzii Sph. gossypii est typicus observa- tionis et auctoris et Friesii." (9) My own examination in July, 1899, of material preserved in the de Schweinitz herbarium in Academy of Natural Sciences of Phila- delphia, led to no different result. The specimens were examined both by reflected and transmitted light, with a hand lens and with the com- pound microscope. The specimen in the collection proper (books of Sphterias) is labeled "Sphferia gossypii L. v. S. and Fr. Salem." This pocket contained nothing but some dust particles, insect detritus, and fragments, the largest of which was less than 1 mm. in area. These dust partitles and fragments showed no perithecia or Fusarium spores. No fungus with necks either long or short was to be seen. Traces of an unknown, col- orless, very delicate mycelium were observed, and one Macrosporium spore. This pocket also contained the skin of a museum pest. In a separate package labeled "Fungi | Cryptograms | fr. Dr. Schwei- nitz I to Collins," I found, however, an uninjured, well-preserved speci- men labeled in the handwriting of de Schweinitz " Sphwria Gossypii L. V. S. and Fr. Salem." This pocket contains two fragments of cotton 53 capsule pericarp, eacb measuring about 1 by lA centimeters. Tlie inner membrane is wiiite, longitudinally wrinkled, and sound, or at least bears no Spb;i'riaceous bodies, Fusarium spores, or fungoustlireads. Jnplaces it bears tiny rusty specks, which are dead cells of the membrane con- taining some amorphous brown substance. The dark brown outer membrane is raised into numerous small papilla', quite regularly arranged. These papilhe are barely visible to the naked eye and under a lens magnifying onlj^ two to three times might readily be mistaken for buried perithecia. These are undoubtedly what de Schweinitz saw and named Splucria gossypii, but they do not contain any perithecia. Under a lens magnifying ten diameters they look more doubtful and when examined under the compound microscope (dry and crushed in water) they are observed clearly to be not of fungous origin, at least there are no Nectriaceous or Splncriaceous bodies either on the sur- face or in the depths, and no Fusarium spores or hyphie. Such papilla} are very common on the surface of cotton bolls when shriveling, as every one knows who has seen much of the plant, and very often at at least, they are not of fungous origin. The presence of long necks which come to the surface and pour out gelatin and the statement by de Schweinitz that "scarcely a capsule of cotton is to be seen without this Sphicria" make it reasonably certain that de Schweinitz did not found his description on the infrequent little red perithecia which I have discov^ered. Finally, an examination of some of the scanty material of Sxfha'ria gossyjni which has been j^reserved in herbaria seems to indicate that the species was founded on the iiapil- late appearance of the dry cotton capsule and on the fact that the sur- face of cotton capsules frequently exhibits a purple stain and a shining appearance. All the rest of the descriiition (much more of it than I at first supposed) is pure inference — i. e., an easy method of accounting for the i^apillte and the purplish glaze. DESCRIPTION OF PLATE I. 1. Mature perithecia resting on fragment of hypocotyl of an old watermelon plant killed by the internal fungus. Monetta, S. C, September 9, 1895. This figure will answer equally well as an illustration of the perithecia occurring on cotton or cowpea. 2. Immature ascus, with two paraphyses and a few cells of the hypothecium, crushed out of a perithecium on watermelon. The smooth and still uncolored spores are surrounded by granular periplasm. Monetta, S. C, September 9, 1895. 3. Mature ascus taken from a perithecium which grew on mycelium derived from the ascospore shown in G. The periplasm has disappeared and the spores are now brown and have a thick wrinkled epispore. 4. A group of immature asci crushed out of a perithecium on cowpea. 5. Ripe ascospores of the watermelon fungus highly magnified. 6. Germinating ascospore, cowpea fungus. Agar-plate culture. James Island, S. C, August 22, 1895. 7. Conidia-bearing mycelium developed from the ascospore shown in 6'. S. Three of the same conidia more highly magnified, one germinating. These conidia are identical in appearance with those produced inside of the vascular sys- tem of the still living stems. (See Plate II, 11.) 9. Surface or dry-air conidia of the cowpea fungus. These were taken from the conidia beds shown in 11 and 13. 10. Fragment from a cross section of the living cowpea stem, about 2 feet from the ground, showing a group of vessels infested by the fungus (c, cambium; x, xylem; p, pith), imbedded in paraffin, cut on the microtome, and stained for many hours in acid haematoxylin. This picture represents about one-thirtieth of the vascular ring, nearly all of which was occupied in the same manner. The cortical parenchyma and the bulky pith were still free from the fungus, as is always the case in this stage of the disease. Later the fungus pushes through to the surface, as shown in 11 and 12. James Island, S. C, August, 1895. 11. Surface of dead stem of cowpea, showing rows of conidia beds. The vessels of such plants are always previously occupied by the internal fungus, as shown in 10. The row-like arrangement of the conidia beds is due to the fact that the fungus comes to the surface along lines of least resistance— i. e., through parallel rows of parenchyma cells separating the strands of tough bast fibers (stereome). x2. James Island, S. C, August, 1895. 12. A similar fragment of dead stem of the cowpea more highly magnified. X 10. 13. A tube of boiled rice overgrown by the mycelium of the watermelon fungus (Ftisarium niveum). Culture No. 4, October 8, 1895. Painted October 28, 1895. This culture was bright blue on the start, 2 c. c. of violet litmus water and a few drops of a saturated solution of sodium carbonate (t. 25° C.) having been added to it. The same brilliant colors may be obtained, however, as already stated in the text, without use of litmus — i. e., by simply cultivating the fungus for a few weeks on rice boiled in distilled water — and this figure will answer equally well for such cultures. High up on the walls of the tube (above the rice) the fungus is white. 14. An incipient perithecium developing on mycelium produced by the ascospore shown in 6. With exception of 10, 11, and 12, which were painted under my supervision by Mr. John L. Ridgway directly from the specimens, and 13, which was j)ainted in the same way by Miss D. G. Passmore, the figures are from my camera drawings. 1 was trans- ferred to the plate and painted by Miss Passmore. The rest of the work was done by Mr. Ridgway. Where no measurements are given they are included in the plates or may be learned from the text or from other figures of the same sort in which they aie given. 54 iULL 17. DIV.VE6.PHYS I'l. ^ Ertviii E Sniilli, D.G.fassHiore aitd Jotui L.Ridi^wav. A.tloen K' Co. Lith. Tt.ihiiaoiv. TvEOCOSMOSPORA NOV. GEN. Erwik V. Smith . DESCRIPTIOX OF PLATE II. 1. Closed stoma. Peritheciiim of the cowpea fungus. Diameter of stoma proper, 45 n. James Island, S. C, August 29, 1895. 2. Open stoma of the cowpea fungus, showing periphyses lining the inner wall of the throat. Diameter of the stoma, 30 /<. James Island, S. C, August 29, 1895. 3. Another A'iew of the periphyses, a portion of the inner wall of the neck of the perithecium of the cowpea fungus crushed out and examined in water. James Island, S. C, August 29, 1895. 4. Optical section through perithecium from a dead watermelon stem, showing cavity full of loose ascospores. Some of the asci remained, but the walls of most had dissolved. This figure will answer equally well for the cotton or cowpea fungus in each of which the same phenomenon was observed. Size, 320 by 280 >«. Slightly diagrammatic. Monetta, S. C, September 9, 1895. 5. Peridial cells in the middle or ventral portion of a perithecium of the cowpea fungus. The same are shown less distinctly and somewhat diagrammatically in Plate I, 1, and Plate V, 1. James Island, S. C, August 27, 1895. G. Conidial tufts on the surface of a watermelon stem killed by the internal fungus. Monetta, S. C, July 16, 1894. 7. Top of a similar tuft (melon fungus), showing attached and loose spores in various stages of growth and septation. From a plant destroyed by soil infection. Washington, D. C, September 27, 1894. 8. Parenchyma cells from the living hypocotyl of a young watermelon, third day of the wilt. Parenchyma partially occupied by conidia-bearing mycelium. No fungus on the surface. In this cell there were 26 conidia. The section was jarred repeatedly and somewhat roughly, and finally turned over without disturbing the fungus. Undisturbed nonparasitized parenchyma cells were above and below it, and the conidia were plainly inside of the cell here figured, which was from tissue near a luugus-infested bundle. Hothouse experiment, Washington, D. C, April 10, 1895. 9. An external, lunulate, 3-8eiitate conidium of the watermelon fungus after seven- teen hours in acid cucumber agar. Twenty-four hours later numerous elliptical conidia, like those shown in Plate I, 7, were abstricted. 10. Hypha end, showing abstriction of the internal conidia from mycelium of 5, forty eight hours after the latter was drawn. 11. Internal conidia of cowpea fungus. These spores were taken from the vessels of cowpea stems on James Island, S. C, August 7, 1895. Twenty-nine internal conidia were measured that day, the size varying as follows: Length, 4.5 to 18yu; breadth 2.3 to 4 /i. Figures transferred to the plate by Mr. Williams Welch, from camera drawings by the author. 56 Bull, I 7, Div Vfg Phiys. ci Pa'>h., U. S. Dep1. of Agricultu- Plate II. Fungus of the Melon and Cowpea Wilt. DESCRIPTION OF PLATE III. 1. Microconidium from the interior of a melon stem, germinating after 7 hours in agar. 2. Th« same conidium, after 24 hours in agar, showing numerous branches and. the abstriction of new conidia. Under favorable conditions conidia are formed very rapidly. Three hours after this drawing was made this mycelium had given rise to 40 free conidia and 30 more were in various stages of growth. 3. Formation of microconidia of the watermelon fungus in an agar plate culture. This hyplia end was under continuous observation for two hours and eighteen min- utes, during which time two spores were developed and abstricted and another begun. Room temperature, 27° C. The formation of conidia was carefully followed in a number of other cases. Under the most favorable conditions of temperature and food supply only forty-five minutes intervened between the pushing off of one conidium and the formation and abscission of another. Usually, however, fifty-five to sixty minutes was required. 4. Mycelium and conidia of the melon fungus from a young agar culture. This was one piece of mycelium, broken in the drawing at the place marked x for con- venience of representation on the plate. Mycelium and conidia vacuolate, all formed in the agar, only sterile hyphie ends projecting into the air in this early stage of growth. Two spores germinating. This figure well illustrates the waythehypha end throws oif a few conidia, passes into a vegetative condition, and elongates for a time, with formation of septa, and then once more ceases to elongate and becomes sporiferous. Monetta, S. C, July 4, 1894. 5. Microconidia of the melon fungus from an agar plate, showing the variability in size. Two spores germinating. 6. Conidia and torulose mycelium from a culture of the internal watermelon fungus {Fusariuni niveian) 27 days old. 7. Fragment of mycelium from the same culture as 6. 8. Fragment of mycelium from a culture of the internal cotton fungus (Fiisarium vasinfectum) 25 days old. For comparison with 7. 9. Mycelium and microconidia of the cotton fungus {Fusarium vasinfectum), culti- vated from the interior of a diseased cotton stem received from western Georgia. From a pure white culture 25 days old. For comparison with 3 and 4. 10. Conidia of the cotton fungus from a culture 25 days old, derived from the internal or microconidia. For comparison with 11. 11. Macroconidia and microconidia of the watermelon fungus from a pure cul- ture 5i mouths old, on sterilized horse dung. The mycelium which bore these spores was derived from a spore of the size and shape of the largest here shown. 12. Chlamydospores of the melon fungus. Several germinating. From a pure culture 5i months old, on horse Jung. This culture was derived from a lunulate, ^everal-septate, external conidium. In mass these chlamydospores were brick red, and their contents was considerably denser than has been indicated by the engraver. At 11 a. m., when the examinations began (in water), there were no germinations; at 4.30 p. m. there were many. In the same tube with these chlamydospores were conidia of all the sizes shown in 11, the small spores being much more numerous than the large ones. Figures drawn by the author and engraved on wood by L. S. Williams. 58 Bull, 17, Div. Veg, Pnys. & Patn., U. S. Dept. of Agriculture. Plate III. II 4 3 nSO 12: IS 12-55 Fungus of the Melon and Cotton Wilt. ( From culture nierlia. ) DESCRIPTION OF PLATE IV. Cross section of a mature watermelon stem, showing how the vessels are plugged by the fungus. In this stage of the disease the foliage has suddenly wilted (as shown in PI. VII, 1), but is not yet shriveled; the stem is green and turgid and its parenchyma is not yet invaded. From the surface inward the tissues are as follows : (1) Epidermis, (2) coUeuchyma, (3) cortical parenchyma, (4) bast fibers (stereome), (5) medullary system, (6) ten bicollateral bundles in two rows, (7) pith. The struc- ture of the bundles is as follows : (1) outerphloem, (2) cambium, (3) xylem, consisting of pitted and reticulated vessels held together by wood parenchyma, (4) spiral vessels lying in non-lignified living parenchyma (the primary vessel parenchyma of Strasburger), (5) pseudocambial layer, (6) inner phloem, composed like the outer phloem of sieve tubes and companion cells. On the outer side of the outer phloem may be seen the collapsed remnants of the primary sieve tubes. The middle portion of the stem is occupied by fissures. Section embedded in paraffin, cut on the microtome, and stained in hsematoxylin. Reduced one-third from a pen drawing made directly from the section by Mr. W. Scholl. Diameter of the stem, 4 millimeters. 60 Bull. 17, Div. Veg. Phys. & Path., Dept. of Agriculture w. S( The Wat Cross section of a stem, showing h Plate IV. 1ELON Wilt. he vessels are plugged by the fungus. DESCRIPTIOX OF PLATE Y. 1. Perithecium and ascospores from upland cotton. Salters Depot, Williamsburg Co., S. C, Oct. 8, 1895. Size 352 by 272 /v. For comparison with PI. 1, 1, which, how- t'ver, is less highly magnified. The ascospores of this perithecium were 10 by 10// with a wrinkled esospore. Others on the same specimens were 9 by 9/^ More rarely they were 9 by 10 /< or 9 by 11 //. 2. Eipe ascus and paraphysis from the same lot of specimens as 1. A paraphysis from a perithecium on cowpea was 18 n broad (cell next to the end cell). S. Macroconidia from conidia beds on the surface of killed stems of sea-island cotton. James Island, S. C, June 29, 1895. All transitions between a and h were observed. 4. Internal or uiicroconidia from a diseased okra plant. James Island, S. C, July 25, 1895. Size 8 to 16 by 2.5 to 3.5//. This okra had been planted in place of cotton which wilted and died early in the growing season. Many plants were affected. The stems were 2 feet tall and about 1 inch in diameter at the base. The wood was much browned the whole length of the stem. 5. Macroconidia from external conidia beds on dead okra plants. James Island, S. C, July 25, 1895. The mature spores measured 27 to 42 by 3 to 5 //. Occasionally one was 5-septate. The vessels and parenchyma of these plants contained a great amount of mycelium bearing such 8i)ores as are shown in 4. 6. Highly magnified cross section of a pitted vessel which is beginning to be occu- pied by the melon fungus. Stem of watermelou. Monetta, S. C, June 26, 1894. Camera drawing from a fresh section examined in water. A great many of the ves- sels were plugged solid by the fungus, as if stuffed with cotton. For the location of these fungus-infested vessels with reference to other parts of the stem see Plate IV. Figures transferred to the plate by Mr. Williams Welch from camera drawings by the author. 62 Bull. 1 7, DU. Veg Phys £c Path., U. S. Dept of Agriculture. Plate V. Fungus of the Cotton, Okra, and Melon Wilt. DESCEIPTIOX OF PLATE VI. Healthy melon vines. Mouetta, S. C. June 28, 1894. For comparison with the photographs shown on the following plate, which were made at the same time and from the same field. The vine in the foreground had about 200 leaves and a melon three-fourths grown. 64 Bull. 17, Div. Veg. Phys. & Path., U. S. Dept. of Agriculture Plate VI. 4133— :No. 17 5 65 DESCRIPTION OF PLATE VII. Vines attacked by the melon fungus. Monetta, S. C. Photographs a. n\. of June 28, 1894. 1. On the afternoon of June 26, when last examined, this vine was to all appear- ances as healthy as that shown on Plate VI. .?. A vine which has been wilted for several days. Healthy vines in the background. 66 Bull. 17, Div. Veg. Phys. & Pam. U. S. Dept of AgticuHure. Plate Vll. y m 'li < t Z — m s: rn Pi 11 o o <: ^-^ -^ H ^ 1 Tq m iq fn s, > 5 m = > — en -I m DESCRIPTION OF PLATE VIII. Melon wilt. The result of soil inoculations. Hothouse experiment. Washing- ton, D. C. Photographed April 17, 1895. Soil inoculated in 1894 with internal melon fungus brought from South Carolina. 1. Two healthy and 3 diseased plants; S. One healthy and 3 diseased plants. In this stage of growth the first symptom is a drooping of the cotyledons ; this is followed by flaccidity of the plumule and a bowing over of the hypocotyl. The parts above ground and the roots also are sound externally— i. e. they are not wounded, rotted, softened, shriveled, or browned in any way. In this early stage of the dis- ease there is little or no fungus in the hypocotyl, but the vessels of the taproot are plugged. ' After photographing, water was withheld from these plants, and as they were iu a dry room they soon died, the healthy ones included. On April 24 each of the 6 plants which were wilted when photographed bore the conidia beds of the external Fusarium, and a further examination showed the bundles and parenchy- matic tissues of these plants to be full of the internal mycelium and microconidia. On the contrary, the 3 plants which were healthy when photographed contained no internal fungus, and there were no Fusarium beds on the surface, although the plants were under a bell jar in moist air for a day or two prior to the examination. ' June 29, 1894, about 1,500 hills of watermelons were planted by the writer on a sandy field in South Carolina, which had been infected from stable manure, and on which most of the melon vines had been destroyed by this fungus in May and June of that year. The disease began soon after the plants came up, and in 6 weeks nearly all of the young melons had wilted, altogether perhaps eight or ten thousand plants. The cotyledons first became flabby and drooped, the first true leaf then wilted, the hypo- cotyl lost turgor and bowed over to the ground, and the plants finally shriveled — i. e., the symptoms were precisely the same as those subsequently obtaiued in Washington with pure cultures of the melon fungus. A hundred or more of these wilting plants were examined microscopically, and in each one the fungus was found in the vessels of the hypocotyl or taproot or both in quantity sufficient to account for the disease and commonly nothing else was present. In many plants which were not pulled until the second or third day of the wilt, the fungus was found pushing out into the parenchyma cells and fruiting therein, as shown on PI. II, 5. July 14 the writer removed 14 healthy-looking young melon plants from as many difierent hills in this field and examined them microscopically for the presence of the fungus. Vines had recently wilted in each of tliese hills. The big seedlings were growing rapidly and appeared to be perfectly healthy above ground and below. In 12 of these plants no fungus could be found. In 1 there was an abundance of the fungus in the big central duct of the taproot and in some of the smaller vessels, but none could be found in the hypocotyl. In the other, there was also no fungus in the hypo- cotyl save doubtfully a thread or two in one vessel, but there was plenty of it in the big ducts of the taproot about 1 centimeter below the crown. No hyph?e threads were observed in the i^areuchyma cells of the roots, which were white and appeai-ed to be entirely sound. These two plants would have wilted in a day or two, and the history of the experiment shows that the other 12 would subsequently have contracted the disease. 68 Bull, 17, Div, Veg. Phys & Path.. U. S Deot of Agriculture Plate VIII DESCRIPTION OF PLATE IX. Watermelon wilt. Hothouse experiment. Washiugtou, D. 0. Spring of 1895. 1. Pots of the same series as PI. VIII, but ijhotographeil some weeks later. All of the plants were killed by the fuugus except the three mentioned above. Seeds planted March 12; photograph made some time in May. 2. Control plants growing in nninoculated (healthy) soil. None of these plants contracted the disease, but toward the close of the experiment all were dwarfed by leaf aphides. 70 Bull 17, Div. Veg. Pnys. Sc Path., U. S. Dept of Agriculture Plate IX. DESCRIPTION OF PLATE X. Melon wilt. Result of soil inoculations ivith the external fungus. Hothouse experi- ment. Washington, D. C. Photographed June 10, 1895. The fungus used for infec- tion was derived from a single, large, lunulate, several-septate spore, taken from the surface of a plant killed by the internal fungus. When photographed there were no conidia beds on the surface of these wilted plants, but in each case the internal fungus was found plugging the vessels of the taproot; the internal conidia were elliptical or bluntly i)ointed, straight or slightly curved, nou-septate, and 8 to 20 by 3.5 to t //. On June 8 the plant at the right was as healthy looking as its companion, but the plant in the middle pot and at the extreme left were already wilted. The three healthy vines afterwards contracted the disease. Bull. 17, Div. Veg Phys. & Path., U. S. Oept. of Agrirulture. Plate X. o o r. 5. ^ i o 3 ■d Bulletin No. 18. v. P. P.-68. U. S. DEPARTMENT OF AGRICULTURE. DIVISION OF VEGETABLE PHYSIOLOGY AND PATHOLOGY. B. T. GALLOWAY, CHIEF. THE PHYSIOLOGICAL ROLE OF MINERAL NUTRIENTS. BY OSCAR LOEW, OF THK Divisiofi of Vegetable Physiology and Pathology . WASHINGTON: GOVERNMENT PRINTING OFFICE. 1899. LETTER OF TRANSMITTAL. IT. S. Depart:vient of Agriculture, Division of Vegetable Physiology and Pathology, Washington, D. C, September 28, 1899. Sir: I respectfully transmit herewith a bulletin prepared by Dr. Oscar Loew, of this Division, on the physiological role of mineral nutri- ents. For several years Dr. Loew has been engaged in the study of the functions of mineral nutrients in plants. This work has led iuto new fields, and has resulted in discoveries which it is believed will be of much practical value to agriculture. The importance of mineral nutrients is so well known as to need little comment, but the part each plays in the life of an organism is still largely a matter of conjecture. It is evident that we can not hope to understand nutrition until we become better acquainted with the physiological action of the nutrients themselves. The main object of this bulletin is to show what has been accomplished in this direction, and to encourage and stimulate work along the lines laid down. The matter has been prepared primarily for teachers and experiment station workers, and is therefore treated from a technical rather than a practical standpoint. I respectfully recommend that it be published as Bulletin No. 18 of this Division. Respectfully, B. T. Galloway, Hon. James Wilson, Chief of Division. Secretary of Agriculture. CONTENTS. Page. General Remarks ox the Mineral Constituents Found in Organisms. 5 Historical notes 5 Ecological and physiological rule of mineral substances 8 Mineral componnds found in organisms 9 General value of certain mineral salts 9 The low atomic weight of the mineral nutrients 11 The Physiological R Ale ok Phosphoric Acid 11 Relation of phosphoric acid to proteids and to the division of cells 11 The physiological importance of lecithin 12 Phosphoric acid in chlorophyll 14 Potassium phosphate as a cell constituent 14 The Ph\ siological R6le of Iron Compounds 15 Relation between the coloring matter of the blood and of the leaf 15 Influence of iron and other mineral nutrients on the formation of chloro- phyll 15 Fertilizing effect of iron salts 16 Organic compounds containing iron . . 17 Iron in fungi 17 Manganese in plants 17 The Physiological Role of Halogen Compounds 18 Plants raised without chlorids 18 Value of potassium chlorid for buckwheat 18 Injurious effects of chlorids on plants 19 Absorption of chlorids by aquatic plants 19 Sodium chlorid in animals 20 Physiological occurrence of calcium fluorid 20 Behavior of plants to potassium brouiid 20 Relations of organisms to iodine compounds 20 The Physiological Role of Alkali Salts 21 Importance of potassium for the formation of starch and protein 21 Beneficial action of sodium salts 23 Can potassium salts be reitlaced by rubidium salts in green plants and in animals 1 23 Necessity of sodium salts for animals 25 Behavior of fungi toward rubidium salts 25 Physiological superiority of potassium salts . , 26 The Physiological Role of Calcium and Magnesium Salts 28 Distribution of lime and magnesia in plants 2S The physiological importance of lime salts in plants 30 Views on the functions of lime salts 33 Formation of lime incrustations 38 Can calcium in plant cells be replaced by strontium? 39 Poisonous action of magnesium salts 42 3 The Physiological K6le of Calcium and Magnesium Salts— Cont'd. Page. Life without lime salts 44 On possible relations between the lime and the transportation of starch . . 46 The physiological role of magnesium salts 47 Increase of magnesia in oily seeds 50 Necessity of magnesium salts for fungi 51 Can magnesium salts be replaced by beryllium salts ? 52 Importance of lime salts for auimals 55 Proportions of lime and magnesia in animal organisms 56 Behavior of animals to strontium salts and oxalates 58 Final remarks 59 THE PIIVSIOLOOICAL ROLE OF MINERAL NUTRIENTS. By OSCAK LOEW. GENERAL REMARKS ON THE MINERAL CONSTITUENTS FOUND IN ORGANISMS. HISTORICAL NOTES. The functions of the mineral nutrients in plants and animals consti- tute a highly important problem. Normal development is im])eded by tlie decrease and entirely prevented by the absence of even a single nutrient, and gradual decline, disease, and finally death will result from the continued withholding of any such substance. Thus yellow spots will develop on the leaves of the sugar beet when the soil is defi- cient in lime; mold fungi will not deveh)p spores but only mycelium when the amount of magnesia in the nourishing medium becomes too small; and even the primary segmentation will stop entirely in the fecundated eggs of lower marine animals when lime salts are withheld, while every further physiological action comes to an end as soon as sodium salts are substituted for potassium salts in the cells. Pigeons die in a few weeks when fed with materials too poor in mineral matter, and dogs can not subsist on meat which has been macerated with cold water, by which means most of the mineral matter is removed. The result of eating such food is weakness of the muscles and nervous excitability, which finally lead to death with spasms and the symptoms of suftbcation. In examining highly differentiated plants and animals there are observed not only certain deferences as to the total sum of mineral matter in the different organs of the same organism, but also certain regularities as to the proportions of the mineral constituents. On the other hand, pathological conditions lead to a jDartial excretion of tlie mineral matter. Thus, in tuberculosis of man the excretion of lime and magnesia is increased, and in diabetes the increased excretion of lime is a specific symptom. These facts can be understood properly only when it is admitted that for the normal functions of normal organs a certain amount of lime and magnesia is indisi^ensable. 5 Humphrey Davy ' was the first savant to consider the mineral con- stituents essential for the development of plants. He says: "The chemistry of the simpler manures (the manures which act in very small quantities, such as gypsum, alkalies, and various saline substances), has hitherto been exceedingly obscure. It has been generally supposed that these materials act in the vegetable economy in the same manner as condiments or stimulants in the animal economy, and that they render the common food more nutritive. It seems, however, a much more probable idea that they are actually a part of the true food of plants, and that they supply that kind of matter to the vegetable fiber which is analogous to the bony matter in animal structures." Davy mentions among other things the beneficial action of gypsum, bone dust, and slaked lime. Indeed, the favorable effects of wood ash, bone dust, and liming upon vegetation have been known since olden times. Furthermore, mills for grinding bones existed early in this century in France and England, and enterprising men went so far as to dig up battlefields in Europe and unearth thousands of tons of bones for agri- cultural purposes. SprengeP was the second one to express an opinion on this sub- ject. He says: " We can accept it as an indisputable fact that mineral matters found in plants also are real nutrients for them, and that it is not their action upon the humus which makes them important, since gypsum, potassium sulpliate, and calcium phosphate do not at all act upon the humus.''-' In quite a diflerent sense Berzelius argued the same year that the action of lime is simply that of a stimulant for the plant and a solvent for the humus, while lime and alkali promote the rotting of organic materials, as manure. After Sprengel followed Liebig (1840), whose theories received sub- stantial support in the important researches of Wiegmann and Polstorf (1842). However, great as was Liebig's merit in overthrowiiig the dominant theory of the nourishing qualities of organic matter, called humus, in the soil, and in showing the absolute necessity of mineral salts in plants, the fact can not be denied that he made various errors, especially in his earlier years. For instance, he at one time believed that mineral bases serve merely to neutralize the organic acids in the plant and that they could replace each other, and further, that alkaloids in plants could play the part of mineral bases. He ascribed certain ' Elements of Agricultural Chemistry, Loudon, 1814. ^Theorie der Dlinguug, 1839. ''As a significant fact it may be mentioned that the Prussian Academy of Sciences, in the year 1800 oftered a prize for an investigation to decide whether the mineral matters found in plants are taken up from the soil or whether they are produced in the plants themselves hy vital power. This question was treated hy Schrader, whose decision was in favor of the latter opinion. How much farther advanced was Sanssure, who, in 1804, declared that the mineral matter of humus contribute in a certain degree to its fertility, since the same are found in the ashes of the plants (Recherches sur la vegetation;. diseases of plauts solely to tbe deficiency of mineral matter in the soil, but later investigations liave demonstrated that fungous or animal parasites are the true causes. After about twenty years of hard fight- ing the importance of Liebig's mineral theory was in the main recog- nized and the old humus theory abandoned. However, his opiuion that the mineral bases replace each other has been proved to be erro- neous by the experiments of Wolff, Knop, Hellriegel, and others. The indispensability of potassa was proved, especially by Friedrich Nobbe, Schroeder, and Erdmann (1871), as was also the noxious character of lithium salts for Phanerogams. But the real significance of tlic bases in the plant cells has not beeu cleared up, and, as a botanist has exi)ressed it, the solution of these problems must form an important final goal for every plant physiologist. When Liebig had galled attention to the necessity of certain mineral constituents in plants he set bis assistants and students at work to analyze the ashes of a great number of plants. He published an account of these analyses in his works on agriculture, but a more com- prehensive review on plant ashes is given in the tables of E. Wolff.' These results show that the quantitative composition of tlie ash of one and the same plant varies according to the soil upon which it is grown, but that qualitatively there is no difference. This observation, which led Liebig to erroneous assumptions, was properly explained much later. It was found that every plant absolutely reciuires a certain minimum of each mineral nutrient, and that in most cases besides this minimum it takes up not only an excess of these various compounds, but also substances which are perhaps useful but not absolutely necessary for plant functions, such as sodium salts and silica. In the case of potas- sium or calcium salts a moderate surplus is not noxious. A large excess of lime taken up can be easily excluded from secondary influ- ences by transformation into oxalate or carbonate — salts which are often produced by plants. Plants adapted to saline desert vsoils show incrustations on their leaves, which may sometimes contain, in addition to chlorid, nitrate, and sulphate of sodium, more than 50 per cent of calcium carbonate. The surplus of mineral matter found in plants — nutrient as well as indifferent compounds— depends to a great extent upon the intensity of the current of transpiration, which explains why herbaceous plants show a higher percentage of ash for the dry matter than do the leaves of woody plants. While cabbage leaves, which have about 90 per cent water, contain 15 to 18 per cent ash for the dry matter, the leaves of potatoes, clover, and grass, having 78 to 80 per cent water, contain only G to 9 per cent ash for the dry matter. In trees adapted to moist soil, for instance, Salix, Populus, Acer, and Tilia, the leaves contain more water and also generally more ash for the dry matter than do the leaves of trees in which transpiration goes on more slowly, such as Quercus, Ascheu Analysen (2 volumes), Berlin, 1871 ;ind 1880. 8 Fagus, and the common kinds of Pinus. While leaves of Acer show 7 to 9 per cent of ash and those of Salix 4 to C per cent, the leaves of Pinus montana and P. austriaca show only 0.58 per cent and 0.74 per cent, respectively (Ebermayer). There is more ash in the leaves than in the roots or stems, more in the roots and stems than in the seeds, and more in the seeds than in the wood. ECOLOGICAL AND PHYSIOLOGICAL BOLE OP MINERAL SUBSTANCES. A question of fundamental importance is whether a certain mineral constituent has one or several functions to perform, and in the latter case whether at least one of these several functions may not be per- formed by some other related constituent— in other words, whether a partial substitution in the organisms would be possible. When a mere neutralization of acids or an osmotic action is involved there can be no doubt that potassa or lime may be replaced by soda, or when incrus- tation of a tissue is necessary for protection the place of calcium car- bonate might be taken even by silica. The solution of various mineral salts produces osmotic pressure and motion required also by animals. Thus beef tea containing 0.35 gram of salts per liter, exerts an osmotic energy of several atmospheres, of which, however, only about one-fourth can be realized in the stomach, since the blood itself also contains min- eral salts. However, it suffices to produce an aiiueous current from the blood to the stomach, while in the intestines the current takes the oppo- site direction. 1 Such functions are not specific, however. But in the purely physiological functions of a chemical nature not even a partial substitution is possible, notwithstanding that various assertions have been made to the contrary. The most novel supposition in connection with this idea of substitution, and one very amusing to chemists, is that seriously made in a recent text-book of plant physiology, that is, that on other planets there may exist living organisms in which the carbon of the organic matter is replaced by silicon. In order to furnish a foundation upon which to base a theory of the special functions of the various mineral constituents, separate analyses for each kind of organ are indispensable. In former times entire plants or animals were subjected to incineration and the ash analyzed, but such results were of very restricted value." A distinction may be drawn between ecological and physiological functions. In the former case the mineral compound serves either as a mechanical support of the organic forms, as does for example the calcium phosphate in the case of bones, and probably the silica in grasses, feathers,^ and hair; or it furnishes a protection against noxious intluences from the outside, and against the attacks of enemies,'* as do the needle crystals of calcium 1 Koppe, Therap. Monatshefte, 1897. 2 Thus one author has inferred from his analyses that there is less magnesia in cats and dogs and less potassa in dogs than in rabbits (Zeitschr. Biol., Vol. X, p. 321). 3 The organic silica compound in feathers was recently studied by Drechsel. 4It has been asserted that the siliceous deposit in the bark of Fagits and Acer and in the leaves of various other plants forms a protection against parasitic fungi. 9 oxalate in certain leaves against snails; the incrustations of leaves and entire plants (Chara) by caloiiim carbonate; the lime shells of Fotami- nifera, certain worms, mollusca, and eggs of birds; the silica shells of diatoms; and the secretion of dilute sulphuric acid by certain Gastro- poda, as Dolium, Cassis, etc. Finally, mineral matter may be an object of adaptation, as the salts in sea water is for marine animals. How- ever, in this bulletin the physiological role alone is the subject of discussion. MINERAL COMPOUNDS FOUND IN ORGANISMS. The mineral compounds usually found in living organisms are phosphates, sulphates, carbonates, chlorids, silica, iron compounds, magnesia, lime, soda, and potassa; while in plants nitrates, manganic compounds, and minute quantities of fluorides also often occur. Small quantities of iodine compounds are found in both kingdoms. Bromine compounds occur in sea weeds. Occasionally there are present in plants small (juantities of titanic and boracic acids, lithia, and alumina, and of the oxids of lead, zinc, and copper." Sodium salts are not necessary for physiological uses of plants, but are for those of animals. Calcium salts are of great importance for plants and animals, only the lower fungi and lower alga? being able to do without them. Magnesium and potassium salts, however, can not be dispensed with by any living cell any more than can phosphoric acid. Manganese, which was shown by Eisse to be incapable of replacing the iron in plants and was believed to be entirely useless, forms, according to recent researches of Bertrand, an essential constituent of the vegetable oxidizing euzynis, and hence may be also of physiological interest.- The nitrates and sulphates present in plants serve, in regard to their acids, chiefly as sources of nitrogen and sulphur for protein formation, and consequently do not require further discussion. As physiological elements these must be designated potassium, sodium, calcium, magnesium, iron, phosphorus, chlorine, iodine, carbon, hydrogen, nitrogen, oxygen, and sulphur. GENERAL VALUE OF CERTAIN MINERAL SALTS. Mineral salts have not only to perform ecological as well as specific chemical functions, but also seem to contribute directly to the mainte- nance of the continuance of the living condition of the protoplasm. A most striking instance of this is the rapid dying of infusoria in distilled water. The writer entertained for a time the supposition that this 'Lippmann observed in the sugar beet not only boric acid and copper oxid, but also traces of vanadin and caesium compounds (Bot. Jabrb., 1888). Wait found 0.31 per cent titanic acid in the asli of oak wood, 0.11 per cent in the ash of apples, and traces of it in bones and meat, and Dunnington found it in many soils. ^According to Lepinois, iron can replace manganese in this regard, and the forma- tion of oxidase in plants raised in the absence of manganese was further observed by Albert F. Woods. 10 phenomeiiou migbt be due, as iu the case of the alga Spirogyra, to slight traces of copper sometimes found iu distilled water and derived from the copper vessels used in distilling. Experiments were there- fore repeated, water distilled from glass vessels being used, but the effect was the same — the infusoria died with bloating, their j^rotoplasm swelling and disintegrating. The only conclusion that can be drawn, therefore, is that the distilled water extracts from them traces of neces- sary constituents, which must be of mineral nature, since common water containing some mineral matter has no such action, but forms the very medium of existence for these organisms. A similar effect could not be observed with the same distilled water on algse cells, which may remain alive in it for a considerable time, although the growth ceases. But here the walls of the cytoplasm are probably of greater density, which would prevent the mineral matters of the cell from passing easily to the outside. This phenomenon observed in the case of infusoria strongly resembles that of the red blood corpuscles and leucocytes, which are adapted to the degree of concentration of the serum, and which die when transferred into distilled water, but remain alive for some time in a sodium chlorid solution of O.G per cent. The nature of the mineral salts loosely bound by the proteids of the living matter may vary with the character of these i)roteids. In the one case it may be sodium chlorid, in another the secondary potassium phosphate, and in a third a calcium salt. It should be pointed out once for all that we can hope to understand the living state of protoplasm only when the proteins of the living matter are recognized to be chemically labil bodies, which the slightest influ- ence often suffices to transform into the more stable isomeric forms of dead matter. Relatively stable proteins are also those in milk and the reserve proteins in eggs and seeds. Spontaneous transformations of labil compounds into stable ones by atomic migration often take place very easily, for example, when certain amido aldehydes or amido ketones are liberated from their combination with acids. Years ago M. Nencki ' recognized the importance of the mineral mat- ter combined with the plasma proteids. "All proteins ^ occurring in the living organisms are combined with mineral substances, whereby the proteins concerned acquire specific i)roperties and functional sig- nification iu the organisms." It may be i^roper here to call attention to another phenomenon, first recognized by Wolff. He determined the minimum of each mineral nutrient necessary for the normal development of the oat plant when the other mineral nutrients were in excess, and found that when all the mineral nutrients are offered in the determined minimum amounts at the same time it is impossible for the plants to fiower and fruit 'Arch. des. Sci. Biologiquea de St. Petersburg, 1894, Vol. Ill, p. 312. ^While "protein"' is the general denomination of all kinds of albuminous sub- stances, the word "proteid'" applies especially to the complex proteins, e. g., nucleo-albumin, mucin, hiemoglobiu, etc. 11 normally. When only tbe absolutely necessary minimum of one of the nutrients is otl'ered a certain surplus of some of the others must be present. THE LOW ATOMIC WEKIHT OF THE MINERAL NUTRIENTS. A review of the elements necessary for organic life shows at once that they have low atomic weights, iron, with an atomic weight of 56, having the highest among them. This is due, according to Leo Errera, not only to their more fre([uent occurrence in the various compounds making up the earth's crust, but also to their higher specilic heat. Thus water, constituting as a rule two-thirds to three-fourths and sometimes even more of the weight of a living organism, has the highest specitic heat of all substances, consequently it can diminish the etlects of rap- idly changing temperatures upon life. THE PHYSIOLOGICAL ROLE OF PHOSPHORIC ACID. RELATION OF PHOSPHORIC ACID TO PROTEIDS AND TO THE DIVISION OF CELLS. Phosphoric acid is, above all, necessary for the formation of lecithin' and the nucleoproteids, e.g., chromatin- and plastin— the most essential constituents of the nucleus and plastids. This makes clear the state- ment of former writers, that phosphoric acid "follows the proteids," since every new cell requires them. Wherever phosplioric acid is transformed from the dissolved condition to an insoluble (compound, as in the formation and growtli of the nucleus, fresh quantities must move thither, according to the law of diffusion. The embryos can develop by cell division only when phosphates are stored up in suflBcient quan- tities in the seeds for the formation and increase of the nuclear sub- stance in the new cells. Phosphoric acid, further, is not only contained as calcium and magnesium phosphate in the globoids, but is also dis- tributed in the seeds as dipotassinm phosphate. The observation that the total mass of protein in seeds is increased by an increased supply of phosphoric acid would also be easily under- stood on the basis of the hypothesis of ytrasburger and Sclimit sev- eral years ago made some interesting observations on the migration of magnesia. He found that magnesia as well as nitrogen migrates from the trunk to the points of seed formation, and in a smaller measure also sulphuric and phosphoric acids do the same. The decrease of the magnesia in the wood extended to ninety annual rings. The wood of the tree was analyzed in zones of thirty rings eacli. The percentage of lime and magnesia in the ash are given as follows, as is also for comparison the composition of a beech tree of the same age which had grown near by, but which bore no fruits that year: Lime and magnesia in afruitinij beech and in a control beech. Part of tree. Beech tree in fruit. In the ash- Beech tree not in fruit. In the ash- Lime. Magnesia. Lime. Magnesia. "Rnrlr . Per cent. 85.05 33.92 34.13 35.98 33.36 Per cent. 2.60 12.65 11.95 12.15 13.36 Per cent. 82.10 27.69 31.52 33.55 27.59 31.21 Per cent. 3.65 Zone 1 - 29. 25 26.72 Zone 3 20. 39 7nTip 1 19.02 11.00 Forstl. Naturw. Zeitschr., 1892. 30 As shown by the table, there was relatively a most striking decrease of magnesia to lime in zones 1, 2, and 3 of the trunk of the seed beech as compared witli the corresponding zones from the control beech. The leaves of aquatic plants are also rich in lime. The pro- portions of magnesia and lime were found to be, in Nymplicea lutea, 1 : 8.5 ; in Lemna, 1 : 3.3 to 1 : 7.6 ; and in Elodea canadensis^ 1 : 8.4. Also alg.ie show similar proportions, as seen from the ash analyses of Spirogyra nitida by Pennington (1896) and of fucoids by Godschens (1854). Algjie incrustated with calcium carbonate must, of course, be here excluded. From what has been said under this head it will be seen that the analytical investigations of the ash of plants show (1) that lime and magnesia are present in every part of the plant, and (2) that the leaves contain relatively more lime and the seeds relatively more magnesia than the other parts of the plants. These characteristics can not be accidental, but must be the result of certain functions. THE PHYSIOLOaiCAL IMPORTANCE OF LIMB SALTS IN PLANTS. The more leaf surface is developed in a given time, the more lime is necessary. A normal crop of wheat requires per hectare (nearly 2.5 acres) about 11.6 kilos; sugar beets, 30.2 kilos; grass, 49.4 kilos; clover, 111.8 kilos; and tobacco, 153.7 kilos, while a normal growth of wood needs annually about 20 kilos of lime, besides 7 to 16 kilos of mag- nesia, 2 to 10 kilos of potash, and 0.8 to 4 kilos of phosphoric acid. When the large demand for lime salts by plants is taken into considera- tion, it is easily understood that an absence or deficiency of lime becomes apparent very early. Stohmann' kept maize shoots alive for some time in a culture solution free from lime, but all development gradually ceased with the consumj)tion of the stored-up lime. However, when at the end of several weeks some calcium nitrate was added, a very striking effect was noticed, hardly five hours elapsing before new buds pushed out from the sickly looking tips. Heiden ' observed that maize and peas in culture solutions without lime lived only four weeks, and reached respectively only 18.9 and 27 cm. in height. In culture solutions without magnesia, however, maize lived ten to twelve weeks and peas lived eight weeks and attained a height of 44 and 30 cm. respectively. In solutions without potassa or phosphoric acid, but otherwise complete, such plants lived from eight to twelve weeks. The absence of lime, therefore, was felt first, owing lAnu. Chem. Pharm., Vol. CXXI. -Centralbl. f. Agr. Chem., Vol. XVII, p. 622. Prianisbnikow observed that shoots develop quicker iu a solution of gypsum than in distilled water, which fully accords with the writer's observations. Seedlings of Phaseohis, Fisum, and Ciicurhita kept in distilled water die before all the reserve material is consumed. An addition of a calcium salt to the distilled water leads, however, to the perfect exhaustion of the reserve stores (Boehm, Liebenberg). probably to the relatively small amount of lime in the reserve store of the seeds. Palladin ' placed etiolated leaves of Vicia faba on the surface of dis- tilled water, on a 10 per cent cane-sugar solution, and on solutions of 0.3 per cent calcium nitrate with and without the addition of cane sugar, but a noticeable growth was observed only where both sugar and cal- cium nitrate were present. The same author- has found that etiolated leaves of Vicia faba contain less lime than do green leaves. His analysis showed that there were contained in 1,000 parts of green leaves 13.3 parts of lime, but in 1,000 parts of etiolated leaves only 2.6 parts of lime. The former yielded 10.3 per cent of ash, the latter 7.54 per cent. Stoklasa found in diseased leaves of the sugar beet less than half the amount of lime present in healthy leaves of this i)lant. Church's investigations ' with albino leaves demonstrated that the composition of their ash is very different from the ash of healthy leaves, as the potassa is considerably increased in the white leaves, while on the contrary the lime is more abundant in the green leaves. It is to be regretted, however, that the author did not determine separately the amount of lime present as oxalate and as carbonate and that portion of the lime belonging to the organized matter itself, calculating the results for equal surfaces in both cases. It is also very characteristic that the lime content of the phanerogamic parasite Cuscuta, which forms no chloroplasts in the full-grown state, amounts to only 2 per cent in the ash, while the clover, its host, is very rich in lime. Another interesting case, showing a decrease in lime content in dis- eased leaves, was observed by Dr. Erwin F. Smith in his studies of the peach yellows. He gives the percentage of lime in the ash of the healthy leaves, according to analyses made by Mr. N. E. Knorr, as 40.58, and in the diseased leaves as only 23.88.* According to a later analysis, made by Dr. Eastwood at Dr. Smith's re(iuest,the ash content of healthy twigs of one season's growth is given as 2.10 to 2.58 per cent and that of diseased twigs as only 1.6 per cent, and of healthy twigs from another orchard as 1.4 per cent and of diseased twigs as only 1 per cent.^ In these cases the amount of lime was also less in the diseased iBer. d. Dent. Bot. Ges., 1891, p. 230. 2Ibid, Vol; X, p. 179. 3 Jour. Chem. Soc, 1878 and 1886. The investigations were made with Quercus rubra bearing some albino branches, and also with albino leaves of Plectotjyne varie- gata and of Hedera helix. * Smith, Erwin F., Bull. No. 4, Division of Botany, U. S. Dept. of Agr. ■'^Smith, Erwin F., Bull. No. 4, Division of Vegetable Physiology and Pathology, U. S. Dept. of Agr. 32 leaves, while potassa, magnesia, and in most cases phosphoric acid also were relatively increased, as will be seen from the following table: inalytical data from diseased and healthy trees from four orchards. [Per cents in the ash.] Orchard A, at , Orchard B, at Magnolia, Del. | Dover, Del. Orchard C. at Orchard D, at Still Magnolia, Del. i Pond, Md. Healthy. Dis- eased. Healthy. Dis- eased. Healthy. eSl^. ^^^ti'y- Dis- eased. 40.58 4.81 15.52 7. 55 23.88 61.21 47.61 7.65 20.19 12.63 48. 85 43. 68 40. 54 34.75 Magnesium oxid Potassium oxid Phosphoric acid 5.97 31.86 13.79 5.62 15.02 10.63 3.21 28.26 10.45 4.31 32.51 9.29 2.85 30.18 12.00 10.23 30.76 16.86 The observations which Honda and the writer ^ made with young pine trees cultivated in pure quartz sand moistened with culture solu- tions free from lime have shown that the leaves reached only half their normal size, and that the young trees gradually perished. Bokorny- has cultivated algte {Spirogyra, Zygnema, and Mesocarpus) in culture solutions, in one of which there was no lime, in another no magnesia, and in a third neither lime nor magnesia. These culture solu- tions were kept in aluminium vessels to avoid any trace of substances derived from glass. In the complete solution a normal formation in every respect was noticed. In the solution in which lime was absent the first phenomenon to occur was a decrease of chlorophyll, the chlorophyll band of Spirogyra diminishing not only in breadth and thickness, but also in length, and the original spiral finally becoming a straight line parallel to the longer axis. Some starch, however, was still produced, which proves that it is not the lack of organic matter and of potassa which here brings on this shrinkage, and that the result can be attributed only to the absence of lime. In the solutions in which magnesia and lime magnesia alone were absent, the volume of the nucleus decreased considerably, as well as that of the chlorophyll bodies. The writer has repeatedly observed that Spirogyra majuscida collected from swamj^s containing only traces of lime had very slender chlorophyll bands and scarcely any starch, but that they contained much storedup albumin. When placed in culture solutions contain- ing a moderate amount of lime salts the bands soon became broader. Rudolph Weber ^ instituted a series of experiments with cultures of peas under glass of different colors, and compared these plants with 1 Coll. of Agr., Bull., Vol. II, Xo. 6, Tokyo, Japan. '^Bot. Centralbl., 1895, Xo. 14. The complete solution contained — Per cent. Potassium nitrate 0. 04 Potassium sulphate 03 Mouopotassium phosphate 03 Calcium nitrate - 03 Magnesium sulphate 03 3 Landw. Vers. Stat., 1875, Vol. XVIII, p. 19. 33 plants "Town in very faint li^ht and with normal control plants. The plants were grown in i>nritieroperties decrease rapidly.^ Phanerogams also are easily attacked by oxalates. When i)laced in a 2 per cent solution of neutral potassium oxalate, the nucleus of an onion shows a contraction of about one-fifth of its normal diameter within ten to fifteen minutes. Leaves of Elodea canadensis and VaUisneria sjnraUs were killed completely in thirty-six hours^ in a 1 per cent solution. The control experiments with potas- sium tartrate or sulphate, failed in all cases to show similar action. The claim, therefore, that lime salts are necessary to preci])itate tar- taric acid in plants that contain tartrates instead of oxalates has no sup- port, since neutral tartrates are not poisonous, as are neutral oxalates. Tbe cytoplasm succumbs last, and its death is probably a secondary effect, due to the death of the nucleus and the chlorophyll body. Indeed it can be easily seen that the cytoplasm dies sooner when the number of chlorophyll bodies contained in it is increased. It is on this account that the circulation of the cytoplasm lasts much longer in the root hairs of Chant when under the intluence of a dilute solution (0.5 per cent) of potassium oxalate than it does in the cells of the internodes filled with chlorophyll bodies. An equally dilute solution of neutral potassium tartrate shows no injurious action in the same length of time. The writer's explanation of the poisonous action is as follows: Judging from the most characteristic properties of soluble oxalates, that of precip itating lime from even highly diluted solutions of lime salts and that of depriving lime compounds generally of their lime and of converting it into the insoluble oxalate, he inferred from the peculiar poisonous action the existence of calcium protein compounds in the organized particles from which the nucleus and 'This, however, is not tbe case with free oxalic acid (p. 38). 2 There are some remarkable cases in which monopotassium oxalate exists in the cell sap and still produces no injury, as, for instance, in Humex and Oxalis. In tbese casesitis necessary to assn:i:e an nuusnal density of the tonoplasts— that is, a density sulhcient to protect the nucleus and protoplasm. 37 the chlorophyll bodies are built up. Such ors'iiuized cal- cium couipounds would have a well-dertiied capacity for imbibition, which would change with the replacement of the calcium by another metallic element, and this altered water content must lead to a dis- turbance in the structure, which must ])rove fatal if not remedied in its initial stages. A peculiarity of protoplasm is that alteration of the structure is soon followed b> the chemical change from tlie active to the })assive moditication of its proteids. Now, when potassium oxa- late acts on the inferred calcium protein compounds tliey yield in addition to cal(;ium oxalate the corresponding potassium protein compounds, which, on account of the different capacity for imbibition, can notphysio- logically replace the calcium compound. Moieover, neither tartr.te nor sulphate (which act much less energetically than the oxalate on calcium compounds') atta(;k the nucleus or the chlorophyll bodies. This also shows j)laiuly that it is impossible to accept the view that potassium oxalate becomes dissociated in the cells and that it is the free oxalic acid which, on account of its acidity, kills the nucleus, since potassium nitrate would be expected to act in just the same way.- It will of (fourse be difficult to i)rove microchemically the formation of calcium oxalate in the chlorophyll body or nucleus when ])otassium oxalate is left to act upon them, since the amount of calcium in them is naturally very small, Judging from the great molecular weight of tlie organized proteids with which it would be combined. Moreover the formation of distinct crystals of calcium oxalate would be impeded by the peculiar consistency of the living structures. It was claimed that in view of the highly complicated conditions in the cells the assump- tion of a direct connection between a working cause and an observed pathological result could not be admitted, as the latter might be sim- ply the effect of primarily jjroduced " disturbances of nutrition." However, this claim can not be sustained in the case of the a(;tion of neutral oxalates upon the nuclei, for in the first place this proceeds very rapidly in concentrations of over 1 per cent, and in the second place the processes of metabolism in objects like Spiror/yra proceed very slowly. Further observations by Migula'* deserve to be mentioned here, as they demonstrate that free oxalic acid is among the most poisonous of organic acids. For example, in a solution of 0.004 j)er cent of free • Calcinm tartrate dissolves in about 2,000 parts of water. ^When acting on Spirogi/ra the potassium oxalate seems to pass direct to the nucleus through the plasma strings and not through the tonoplast, but on the other hand when potassium oxalate is contained in the cell sap of certain plants it seems to be confined there by the density of the tonoplast, which also prevents its dii'ect contact with the nucleus in this case. In this connection iiigula observed some interesting facts with Spirogyra kept in well water to which very small quan- tities of organic acids had beeu added. These were gradually oxidized in the cells into oxalic acid of which some was retained as neutral oxalate in the cell sap, and yielded a precipitate of calcium oxalate when placed in a diluted solution of lime salts. 3 The Influence of Dilute Acids on Alga-, (Breslau, 1888). 38 oxalic acid the nucleus of Spirogyra orbicularis was observed to swell up, frequently to six times its normal A-olume, and become turbid and opaque, while the cytoplasm still remained alive for some time. In concentrated solutions the cells die too quickly to show such character- istic symptoms, their death being due chiefly in this case to mere acidity. When some lilaments of Spirogyra majnscida were placed in 500 cc. of a solution of free oxalic acid' in even as high dilution as O.UOOl per cent, the writer observed great injury to some of the threads after live days. In most of the cells the plasma strings were retracted, the nucleus was contracted and rolled to the cell wall, and the sinuate margins of the chlorophyll bands were swollen up and numerous little drops became visible in them.^ A very striking feature was the long-continued per- sistence of the turgor under these conditions, this being due to the cytoplasm remaining alive for a considerable time. In equally diluted solutions of tartaric acid most of the cells were perfectly normal after nine days, which shows that the character of acidity at tliis high dilu- tion exerted merely a secondary influence, and that this alone can not account for the action of the highly diluted oxalic acid. FORMATION OF LIME INCRUSTATIONS. It may not be out of jdace here to say a few words about the forma- tion of incrustations of calcium carbonate on certain aquatic plants, especially Chara — a phenomenon which Pringsheim ' tries to explain on the hypothesis that by assimilation of the dissolved carbonic acid the neutral calcium carbonate is produced from tlie bicarbonate. However, the fact that not every plant growing in the same water and near Chara shows the incrustation must lead to the assumption that either the assimilation is of much greater energy in Chara than in many other plants, or that the surface of this plant is especially adapted for the absorption of the neutral calcium carbonate. Hassack^ advanced another hypothesis, tliat is, that the plants secrete an alkaline carbonate, which decomposes the calcium bicarbon- ate. However, the writer has proved this view to be entirely erro- neous." The reaction with phenol phthalein, which Hassack used is not due to an alkaline carbonate, but to neutral calcium carbonate in a colloidal condition. Even the warming of ordinary water rich in calcium carbonate will produce ephemerally a red color with phenol- l)hthalein. 'Purest water distilled from glass vessels was used for all experimeuts with Spirofjyra. -Considerable swelling of the nucleus took place in a solution of 0.01 per cent oxalic acid. 3 Jahrb. f. Wiss. Bot., Vol. XIX, p. 138. ■'Unters. aus d. Bot. Instit. Tiibiugen, Vol. II, pp. 469-475. 5 Flora, 1893, No. 4. 39 CAN CALCIUM IN PLANT CELLS EE REPLACED BY STRONTIUM! It lias loug been recognized that calcium salts can not be replaced by potassium salts or sodium salts. Were it a well-founded hypothesis that calcium salts serve only for certain phases of metabolism and are not connected with more important properties of the protoplasm itself, then there might be taken a plain chemical view of the matter, that is, that the action of the bivalent elements is often different from that of the monovalent elements. Thus, for example, dextrose yields saccharin^ when treated with lime, but not when treated with potassa (Kiliani); calcium carbamate yields calcium cyanamide ni)on heating, while jiotas- sium carbamate yields potassium cyanate (I)rechsel); barium dibrom- succinate yields monobrommaleic acid on boiling of the aqueous solu- tion, while the sodium salt yields monobrommalic acid. It is certainly not the bivalent character of calcium, however, tliat determines its i)hysiological value, for in that case barium or strontium might fulfill the same office, which is impossible. The inability of barium to do this might be explained by the most characteristic ]»rop- erty of soluble barium salts, which is to precipitate sulphuric acid from even high dilution of sulphates, hence in plants the assimilation of sulphur from sulphates would become an impossibility. However, it would still be difficult to explain why barium salts are i)oisonous ibr animals, and also why strontium salts can not replace calcium salts in either plants or animals.^ The more intimate the connection between the functions of the lime and the vital properties of the cells, the more difficulty will naturally be encountered in an endeavor to substitute strontium for calciun), and experiments made in this connection argue against the possibility of the substitution. The writer made some experiments with an alga {Spirogyra) in 1892 which demonstrated that although this alga can remain healthy for several weeks at the ordinary temperature in a culture solution containing strontium nitrate in place of calcium nitrate, its further growth is nevertheless impeded, and moreover, that there is soon a noxious influence at a higher temi)erature (28° C). Thus, for example, many cells died when kept at 28° C. in a solution of 0.3 per cent stron- tium nitrate, but this was not the case in a 0.3 per cent solution of calcium nitrate. This conclusion has been essentially confirmed by Molisch, who observed the interesting fact that the cell plate in the l^rocess of cell division is not properly formed when strontium salts are pj'esent in place of calcium salts. This occurs even when a small amount of a calcium salt is present, in which case the injurious effects of the strontium salt are not entirely x>revented. The cell plate is the result of the work of the nuclear spindle, and the supposition that the cause of this defective work is attributable to a diseased condition of the nucleus seems justifiable. If the lime were not concerned in the 'This product is uot the sweet saccharin of commerce. '^ Only certain enzym actions form exceptions, as liertrand has shown for pectase. 40 most intimate working of the nucleus the phenomenon in question would hardly be intelligible. Similar experiments with beans and maize were inaugurated later on by Haselhoff,' but he offered calcium and strontium salts together in the beginning and gradually diminished the lime in the culture solu- tion. The plants, however, very probably made use of the occasion to store up a certain amount of lime, which they may have used in the later period, and hence his conclusion that a substitution of calcium for strontium salts is possible can not be admitted. The writer made an experiment with a i)hanerogamous plant also. Branches of Tradescantia, from 12.5 to 12.8 cm. long, were placed in solutions of — Per cent. (1) Calcium nitrate 0.2 (2) Strontium nitrate 2 (3) Calcium and strontium nitrate, each 1 At a temperature of 10-15° C. a decided difference was noticed after twelve days. In the calcium nitrate solution young rootlets 0.5 cm. in length had appeared, but in the strontium nitrate solution only minute knobs were visible. Gradually a difference was also evident between the calcium nitrate and the calcium and strontium nitrate solutions, the root hairs in the former being long and numerous, while in the latter tliey were short and few. However, when the strontium nitrate gradu- ally attained an excess over the calcium salts stored up in the branches, the noxious effect became evident, they having attained a length after forty-two days of only 13 and 13.3 cm., with only two or three leaves on each branch, while those in the solution of calcium nitrate attained :i length of 16, 17.2, and 18 cm., with six to seven leaves on each branch. The leaves of the former branches were partially dying, but those of the latter were still healthy. A control case with distilled water demon- strated beyond a doubt that in the case of the strontium nitrate solu- tion the phenomena mentioned were not merely dne to the absence of the lime, but to a direct noxious action of the strontium salts. The numerous root hairs which developed in the distilled water further justified the conclusion that lime salts were stored up in the stems. Indeed the writer has demonstrated that besides sulphates, the nodes of the Tradescantia stems have stored up in them nitrates, potassium, and magnesium and calcium salts. An undeniable analogy appears to exist, therefore, between the noxious effect of the strontium salts and that of magnesium salts (p. 42), both beginning to be noxious when the amount of lime falls below a certain limit. A series of very instructive experiments were recently carried out by U. Susuki^ with live phanerogamous plants — Mordeum, Fagopyrum esculentum, Phlox paniculata, liubus idwus, and Coreopsis tinctoria. Some of the plants were watered with a normal solution containing 'Landw. Jabib., Vol. XXII, p. 853. ^Bnll. Coll. of Agr., Tokyo, 1899. 41 calciuui in the form of calcium nitrate and others witli sohitions in which the calcium nitrate was replaced by equivalent quantities of strontium nitrate and of barium nitrate. Only the plants in tlie normal solutions showed a stroner mille solution of magnesium sulphate, but remained alive for a long time in corresponding solutions of sodium, potassium, or calcium. In a 1 per cent solution of magnesium nitrate smaller kinds of Spiro- gyra will die in from six to twelve hours, but will live a long time in corresponding solutions of sodium, potassium, and calcium nitrate. Spirogyra which had been kept for several weeks in a healthy condi- tion in a solution of 0.1 per mille of monopotassium phosphate in abso- lutely pure distilled water, died within three to four days when 2 per mille magnesium sulphate was added to this solution, but when dipo- tassium phosphate instead of the monophosphate was used death set in much later, that is, after fifteen to eighteen days. Some threads of Spirogyra majnscula placed in a solution (1 liter) con- taining 0.02 per mille each of magnesium nitrate and ammonium sul- phate, died in from ten to twelve days, while in the control solution, containing calcium nitrate in place of magnesium nitrate, they were still alive after six weeks, although cell division had stopped completely, and the cells exhibited an emaciated appearance owing to the absence of other mineral nutrients. In still another case threads of the same alga were jilaced in a solution of 1 per mille of magnesium nitrate, while 43 in the control case 3 per mille of calcium nitrate was added. ' In the former case death resulted in five days,^ while in tlie latter the cells were still alive after a number of weeks. Lime salts, therefore, are the antidote for magnesium salts. ' Nothing- can replace them successfully in this case, not even nourishment with organic matter.* Microscopi- cal examinations of Spiro(fyra cells exposed to the exclusive action of maunesiiim salts show that the nucleus is attacked first and then the chlorophyll body is injured, the ])henomena closely resembling those produced by potassium oxalate, but while in a 1 per cent solution of magnesium sulphate the nucleus will swell up after twelve hours, in a 0.5 per cent solution of potassium oxalate it will do so in a much shorter time. The noxious action of magnesium salts also soon becomes evident in the roots of seedlings. Thus \lcia and Pisum do not start lateral roots when kept in a solution of 0.5 per cent magnesium sulphate or nitrate, and the root cap and epidermal cells die after a few days. In a solution of calcium nitrate of equal strength, however, development continues. Seedlings of Phaseolus placed in a solution of 0.1 per cent magnesium sulphate, with 0.1 per cent monopotassium phosphate, showed injury to the roots after five days, and the entire plant suc- cumbed soon afterwards. Similar observations had been made by Wolf, by Kaumer and Kellerman, and by others, bnt all failed to recognize the true cause and to ascertain that lime salts alone act as the specific remedy. Raumer^ observed that in Phaseolus multifforvs kept in various cul- ture solutions there was a detrimental effect much sooner when lime alone was absent than when both lime and magnesia were absent. The difference was most striking in the main roots and also in the number and vigor of the lateral roots. Here, then, the noxious effect of magnesia in the absence of lime is again manifested. The writer has made a special study of the development of roots in culture solutions free from lime and from magnesia, using branches of Tradescantia for this i)urpose. These have calcium as well as magne- sium salts stored up in their nodes, and hence some develo])ment of roots is possible even in distilled water, Nevertheless, a most striking difference was noticed, the roots in the culture solutions containing lime but not magnesia producing a "dense forest" of root hairs that reached a length of one-fourth centimeter, while the roots in solutions 1 These observatious the writer desrribefl in Flora, 1892, and also in Landw. Vers. Stat, of the same year. 2 The time is probably prolonged when lime salts are stored up. 3 An addition of strontium salts may delay death for a short period, but it can not prevent it, as do calcium salts. * It may be mentioned that Spirogyra remains alive for fi-om five to six weeks if kept in distilled water. Of course any further development is stopped, bwt assimila- tion and respiration soon reach a suitable eqiiilibrium. " Landw. Vers., 1883, Vol. XXIX, pp. 254 and 268. 44 containing magnesia but no lime, although larger than the others/ pro- duced only a few short hairs. The lack of lime in these roots was felt especially in the epidermis, the interior parts being able to draw a sufficient amount of lime from the stem. Indeed, a microcheraical test showed the presence of lime in the ash of these roots, gypsum needles forming when treated with a little sulphuric acid. The extraordinary effects of lime salts on the development of root hairs is of special interest, as it furnishes the key to the observation of Wolff that the potassium and ammonium salts of the soil are absorbed in increased quantities by plants after manuring with lime salts. LIFE WITHOUT LIME SALTS. While lime salts are indispensable for animals. Phanerogams, and higher algt^, tbey are not so in the case of bacteria, fungi, and lower algae. Thus far no investigations relating to the higher fungi have been made in this regard. The occurrence of lime in the ash of yeast or of tubercle bacilli ^ must be regarded as merely accidental. It was first observed by Adolph Mayer that for yeast magnesia is of greater impor- tance than is lime. Later the writer proved that yeast and bacteria can do without lime entirely, ' and Molisch has observed that this is also true of mold fungi.^ It has been observed, on the other hand, that in cer- tain cases the presence of lime promotes the action of fungi, but this is very probably due only to a secondary effect. Thus, the nitrification in soils is enhanced by calcium carbonate, and, according to Thaxter and Wheeler,' the scab of potatoes and of sugar beets is increased by liming the soil. Recently Laurent" reported that certain bacteria. Bacillus coli communis and B. fluorescens putidus, can attack potatoes in soils which have been strongly limed. He believes that by this means the power of resistance of these plants is diminished so much that the microbes named can commence their parasitic life, and 1 These roots were 4.1 cm. long, while those in culture solutions without magnesia were only 3.2 cm. long. The composition of the complete culture solution in the above case was as follows: Per mill«. Monopotassium phosphate 0- ^ Potassium nitrate • ^ Sodium sulphate K Calcium nitrate o Magnesium sulphate Ferrous sulphate Trace. ^According to de Schweinitz and Dorsett (Centralbl. f. Bakt., No. 23, 1898), the phosphates of sodium, calcium, and magnesium predominate in this ash over that of potassium, while the reverse is true in the ash of yeast. 3 Flora, 1892, pp. 374 and 390. •"Ber. Wien Akad., 1894, Vol. CIII. eStorer, Relation of Agriculture, Vol. II, p. 546. 6Ann. de I'Inatitut Pasteur, 1899, Vol. XIII, p. 1. 45 be fnrtlier asserts tliat only such i)liiiits can resist as have at the same time a great amount of potash and phosi)hoii(' acid.' lioth Molisch- and the writer ' have observed th;it lime is not lequired by the lowe rtbrms of n\g',i\ Molisch proved this in the case of UJolhrix^ Microthamnion, Stivhococcus, and Frotococcus,^ and the writer proved it in the case of a kind of Palmella. Bipartition, zoospores, isogamy, and oogamy represent a scale of progress which probably requires an increasing differentiation of the nuclei. Isogamy in its simpler forms must be distinguished from its more perfected form, as it is found for instance in copulation of Spirixji/ra, where the uniting plasma bodies remain protected by the cellulose wall during the entire process. Some forms of the order Profococ- coidcff multiply only by bipartition, others by swarm spores, certain forms by isogamy, but only two genera { Volvox and Eudorina-') by oogamy. In the order of the Goufervoidecr, Ulothrix multiplies only by isogamy, while (Edofioninm nniltiplies by oogamy also. In other groui>s a still liiger potentialization of the nucleus has to be inferred, as in the Chdracav from the highly ditlerentiated structure. Since neutral potassium oxalate has a i)oisonous effect upon Diatomf, G^dorfonium, Clddophora, and apparently also on Bniparnnldw, the presence of important lime compounds in these oiganisms may be inferred. All these organisms, however, are more differentiated than Ulothrix, which, according to Molisch, can grow in the absence of lime salts. A careful study and comparison of the various chloroplasts of algfe might also show certain advantages in favor of those which reciuire lime for their development. For instance, certain low genera, such as Nostoc and Oscillaria, form no starch, while others do. In such cases starch formation is to be regarded as a step forward, one that depends upon a higher differentiation of the chloroplasts. The beautiful chlo- roplasts of Spirogyra show a high degree of differentiation, the pyre- uoids, wliich form stations in the chloroplasts, being the manufacturers of the starch. It is true Schmitz also observed well-defined chloroplasts multiplying 'A satisfactory explanation as to tlie decrease of power of resistance under the intiuence of such an important nutrient as lime would be very desirable. Perhaps the cells beneath the lenticels are thereby stimulated to growth and open a way for the parasites to enter. ^Sitznngsber. d. Wiener. Akad. d. Wissenschaften, 1895, Vol. CIV. In this article Molisch has also proved that the algic mentioned are incapable of assimilating free nitrogen. This confirms an earlier observation on Noatoc by the writer (Biol. Cen- tralbl., Vol. X, p. 591) and a later observation by Kossowitsch. ^Botan. Centralbl., 1895, No. 52. Probably ^Vosfocat'(w and OsciHa