in the Bromes and their Brown Rust. 
313 
TABLE XXV. 
Summary of results of infection in 843 experiments in which spores 
from B. mollis were used as inoculating material. 
Species infected. 
No. of 
plants. 
Results 
+ in. 
Per cent, 
successful. 
B. erectus 
37 
I 
2, 7 
B. inermis 
3 2 
0 
0 
B. ciliaris (1) 
20 
0 
0 
B. tectorum 
21 
0 
0 
B. sterilis 
90 
4 
4.4 
B. madritensis 
77 
0 
0 
B. maximus 
74 
1 
i-3 
B. secalinus 
61 
3i 
5°*7 
B. velutinus 
62 
35 
56.4 
B. arvensis 
76 
28 
36.8 
B. mollis 
85 
60 
70-6 
B. racemosus 
56 
15 
26-8 
B. commutatus 
45 
H 
3* 1 
B. interruptus 
49 
*9 
38-8 
B. brizaeformis 
18 
5 
27 
B. macrostachys 
12 
4 
33 
B. unioloides 
2 
0 
0 
B. Schraderi (2) 
9 
0 
0 
B. canadensis 
2 
0 
0 
B. giganteus (3) 
1 
0 
0 
B. pratensis (3) 
14 
0 
0 
(1) Species needs examining. (2) Synonym of preceding. 
(3) Species requires further examination. 
I term the ‘infection curve.’ And just as this is possible for 
the case given, so also different curves are obtainable for the 
infections with spores from B. sterilis, B. secalinus , &c . 1 
On drawing these two curves, the result shows there is no 
resemblance between them : nor is there any resemblance 
between the infection curve — as it may be termed — and the 
curves of the sizes or of the number of stomata, of the sizes 
and numbers of hairs (Tables XXVI and XXYII), of the area 
of the leaves (Table XXVIII), or in short the curves of any of 
the structural factors of the leaf. 
In similar comparison of the curves of stomata and of 
infection, it is obvious at a glance that while spores from 
Bromus mollis infect that species readily, as also they do 
1 See Proc. Cambridge Phil. Soc. , 1. c., Tables VI and VII. 
V 
