422 



likely to affect seriously, uuder couditions favorable to its development, 

 an agricultural crop." 



The relation of clover rust to other forms of rust is briefly discussed. 

 '■^Uromyces trifolii has long been known in Europe on many species of 

 clovers and a few other leguminosoB, but has not attracted the attention 

 of mycologists in America until recently. The form on our cultivated 

 clovers was no doubt introduced within a comparatively few years, but 

 that on certain liocky Mountain plants may be native." 



The " rust" i)roper was first reported in America in 1884, by Mr. Hol- 

 way, in the " List of Iowa Uredineic," compiled by Mr. Arthur. 



According to a table given in this article, clover rust has been found 

 in this country from Connecticut to Montana and Utah, and on six 

 species of clover and Glycyrrhiza lepldota. In 1888, according to Pro- 

 fessor Underwood, it was quite injurious to clover near Syracuse, New 

 York. "The uredo and teleutospore stages (the rust proper) were also 

 noticed near Ithaca in 1888, occurring in great abundance. During the 

 year 1889 all three stages have been found on the red and the white 

 clover, the rust stages so abundant as to destroy, it is estimated, 50 per 

 cent of the second or ' rowen' crop. During 1890 it was far less abun- 

 dant. Although in the autumn it was plentiful it did not injure the 

 clover to a great extent." 



The investigations reported in this bulletin were planned by Professor 

 Dudley and carried on in his laboratory, but, as Professor Dudley states 

 in a foot-note to this article, " the discoveries made and the conclusions 

 reached, and in part stated herein, are the result of the faithful, care- 

 ful work of the author, who, as time permits, will follow out the incom- 

 pleted lines of investigation." The fungus in its different stages is 

 described and illustrated, and observations and experiments made at 

 the station are concisely recorded. 



Suvimary. — (1) The parasitic clover rust is chiefly propagated throughout the 

 growing seasou of the host, by the uredospores, owing to their ahundaace, rapid 

 geruiiuation, and the fact that for the most part they reproduce only their own form 

 of spore. 



(2) The germination of the ajcidiospores gives rise to uredo sori, thus demonstrat- 

 ing what has heretofore been merely assumed, viz., the connection of the secidium or 

 " cluster-cup " of clovers with the brown " clover rust." 



(3) Both uredospores and aecidiospores prefer a low temperature in germination. 

 [Experiments showed that the former germinated most rapidly at 11-16° C, and the 

 latter at 15-18° C] This accounts for the fact that during the middle period of the 

 summer of 1890, which was very warm, little of the rust appeared, while more of it 

 was generated during the latter and comparatively cooler months, although at no 

 time did the amount equal that produced during the cool and moist midsummer and 

 autumn of 1889. 



The most important questions remaining to be investigated in the life history of 

 the clover rust are its mode of hibernation and the origin of the aecidia. The two 

 questions are closely linked together, and two theories have occurred as possible 

 solutions of them. First, the secidia may be produced in the spring through the 

 germination of teleutospores which live oyer winter on the dead clover stems. 

 Second, the uredo and teleutospores may gsrminate in the fall, form mycelium in 



