364 
Journal of Agricultural Research 
Vol. II, No. 5 
Table V. Results showing infection produced in Red Amber sorgo by extraseminal 
inoculations —Continued 
Plat 
[Planted at Washington, D, C., in the greenhouse in beds separated by partitions 1 foot deep in the soil; 
counted Apr. 3,1913.] 
Serial No. 
Date. 
Seed lot. 
Spore lot. 
Method. 
Total 
number of 
plants. 
Infection. 
1 
Nov. 1-9 
V 
B and C 
Various b 
67 
Per cent. 
55. 2 
2 
do 
V 
a, control 
16 
0 
2 
do 
V . 
do 
1 16 
! 
3 1 * 2 5 
0 Inoculations by the author. 
6 These inoculations were performed with various methods and stages of growth in an effort to get more 
exact information. With the small number of plants, necessitated by the use of a greenhouse, differences 
in the amount of infection appearing were of little significance in view of the impossibility of properly 
controlling conditions. Most of the plants were not directly watered, except at planting (nor were they, in 
the control), until mature in the spring. Although all were grown in separate beds instead of pots and 
obtained ample moisture from below, they were much stunted by greenhouse conditions. 
« The same plants as above, but counted Oct. 3, 1913, 
While most of the results of these inoculations are positive beyond a 
doubt, an important negative result, as yet unexplained, should be 
pointed out. The conidia, in spite of the care taken to be certain of their 
identity (see p. 341), have never produced the slightest evidence of infective 
power in the few trials made in the field (plat C, Nos. 1,2, and 9; plat 
D, Nos. 2, 3, and 9; and plat E, Nos. 2 and 6). Brefeld (1895, p. 30) has 
found that oat smut, like many other pathogenic organisms, loses its 
virulence after several months in artificial cultures. Unless Sorosporium 
reilianum , as cultivated on carrot agar in these investigations, lost its 
infective power very quickly, however, this explanation does not seem 
adequate, for new cultures grown artificially for only two or three weeks 
produced no infection when inoculated on 15 plants at the same time 
and under the same conditions as plat F, No. 1. The conidia have not 
been observed to produce infection threads as figured by Brefeld (1883, 
pi. 11, 7). 
The first question which arises on considering the fact, here now clearly 
shown, that extraseminal infection does take place, is, What factor has 
been introduced to bring about successful infection when so many former 
attempts had failed? The results given in Table V, while not ex¬ 
haustive, do make clear several of the essential points in the parasite's 
life history which will at least partially answer this question. The 
method designated under the abbreviation bmx will be observed to 
have produced the most consistently positive results wherever tried. 
Except at Amarillo, the only other methods which produced over 20 per 
cent infection were bn(x) in pots in the greenhouses of the Department 
(plat B, Nos. 1 and 2 of Table V), amx at St. Paul (plat E, No. 4), 
and bm(y) at St. Paul (plat E, No. 8), besides the inoculations later 
attempted in the greenhouse (plat F). 
