284 
Journal of Agricultural Research 
Vol. XXXI, No. 3 
cepulae. The diameter of the entire spore ball of U. magica was 
22.19 /x, while that of 27. cepulae was 16.15 /x. All dimensions are 
somewhat greater when the spores are measured in a potassium 
hydroxide solution or when free. When the spores have been kept 
in a dry state for a long time, the accessory cells collapse but become 
distended again when treated with a weak solution of potassium 
hydroxide. Farlow (5, p. 114 ) considered U. magica as identical 
with U. cepulae. Schroeter and Winter (Die Pilze, p. 121, 1884) 
agree with Farlow and unite all the forms discussed here under 
Urocystis colchici Schlecht. Thaxter (If p. 144) and Clinton 
(2, p. 451), on the other hand, point to the differences in the size of 
the spores as sufficient reason for considering the two species distinct. 
If by any chance the host plant of U. magica was really A. nigrum , 
the fact that the writer was unable to cause infection except on the 
tiny cotyledons, while in the exsiccati the fungus is on large leaves, 
seems to the writer another argument against the identity of the two. 
The constant difference in the size of the spores as observed by all 
investigators, and the difference in host plants, seem to be sufficient 
reasons for regarding TJ. cepulae as a species distinct from TJ. magica . 
Another species of Urocystis has been reported on Allium rotundum 
in Europe by various investigators, and has usually been referred to 
TJ. colchici Schlecht., a common species there, occurring on a long 
list of Lilliaceae. Some authors give to the form which occurs on 
Allium the rank of a variety, forma Allii under TJ. colchici because 
of certain morphological differences. The writer made a study of a 
specimen of forma Allii (Fiickel Fungi rhenani No. 2217) collected 
by Fiickel in Austria on Allium rotundum. All characters and 
dimensions were so nearly identical with those of TJ. magica , which 
were studied at the same time, that, from a morphological standpoint, 
no reason could be found for considering the two as distinct. These 
two forms are also doubtfully united by Liro (6) in his recent excellent 
monograph of the genus Tuburcinia. 
After the variety on Allium rotundum has been removed from 
Urocystis colchici, the species as it occurs on other hosts is easily 
distinguished from U. cepulae because the spore balls commonly are 
attached in glomerules of 3 to 5 or more, and a single spore ball may 
have 2 to 3 fertile cells. All dimensions are also larger than those 
of 27. cepulae. 
Rostrup (9, p. 153) found a species of Urocystis on Allium ascaloni - 
cum in Denmark, in 1890, and referred it to 27. cepulae. Liro, how¬ 
ever (i 6, p. 50), refers it to his newly described species U. ferruginea 
which differs sharply from U. cepulae in the red brown color of the 
spore powder, and in the angular spores attached in groups of six 
or more. The writer has not had an opportunity to examine this 
species, but, judging from the description given by Liro, there would 
seem to be no question as to its distinctness and no occasion for con¬ 
fusing it with U. cepulae . 
In 1911, Schellenberg (10, p. 14). described the species Urocystis 
allii from a specimen collected in Switzerland in 1902 on Allium sub- 
hirsutum ana distributed in von Thuemen ’s Mycotheca universalis 
as No. 1219. Although in morphological characters it seems ■ tp 
resemble U. cepulae even more closely than does U. magica, bqtb 
Schellenberg and Liro consider it as distinct. As previously men¬ 
tioned in this paper, A. subhirsutum L. was found by the writer to be 
entirely immune to U. cepulae, which is regarded as evidence con- 
