312 Ward . — On Relations between Host and Parasite 
Columns 2 and 4 are useful checks on column 5, as in the 
case of the stomata already referred to. Also as before, the 
numbers in column 3 have only a relative value, as establishing 
the general fact that the number of hairs on the lower 
surface is greater than that on the upper, thus reversing the 
rule for the stomata. There are one or two discrepancies, but 
I am inclined to regard the results given in column 5 as most 
reliable. 
16. IS THERE ANY RELATION BETWEEN INFECTION 
and the Visible Structural Features? 
It now remains to see whether any of the structural 
peculiarities referred to stand in relation to the comparative 
immunity from the susceptibility of the species or varieties 
of Bromus examined. In order to test this, I adopted the 
following argument and method. 
If the volume of the chlorophyll-tissue, for instance, is a 
factor of importance, then the curve obtained by joining the 
apices of ordinates of lengths proportional to the volumes of 
the chlorophyll-tissue in the various species, ought to bear 
some definite relation to the curve obtained by joining ordi- 
nates proportional to the percentage of successful infections. 
In order to understand the nature of this ‘ infection-curve,’ 
it is necessary to refer to my previous summary of results 1 , 
from which I take the part referring to B. mollis (Table XXV). 
It will be seen that with spores from B. mollis seedlings of 
the same species ( B . mollis) were successfully infected 60 out 
of 85 times — i. e. 70 -6 per cent. ; whereas B. secalinus was 
infected only 31 out of 61 times — i. e. 50-7 per cent., and 
B. maximus only once out of 74 times — i. e. 1*3 per cent., and 
so on. It will be noticed how the successful infections with 
these spores reared on B. mollis all group in the Serrafalcus 
sub-genus, the few exceptions (B. sterilis, B. erectus) not 
seriously interfering with this statement. 
Now, if we join the apices of ordinates proportional to 
these percentages, we have a graphic representation which 
1 Proc. Cambridge Philos. Soc., vol. xi, Part V, pp. 323-328. 
