“ Predisposition ” and “ Immunity ” in Plants. 319 
i.e. the first green leaf, — grown under the same conditions in every 
case, and several methods were adopted for obtaining the measure- 
ments, so that one should check the other. For instance, in count- 
ing the stomata and hairs it will be noticed that I not only obtained 
averages as described below, but in the case of every species I also 
plotted a chart of a square millimetre and so obtained a direct 
count (see Table IV.), and other checks are described in the 
sequel. 
It is of course impossible to suppose that the numbers ob- 
tained can be absolutely correct, but the tabulated results show 
such agreements amongst themselves, for the most part, that I am 
disposed to regard them as more accurate than is usually the case 
in such approximate determinations. It is unnecessary to dwell 
on the laborious nature of the determinations, but of course this 
limits the number of countings. 
I append an account of that part of the analysis which concerns 
the stomata, and in Table V. the summary of results so far as the 
stomata are concerned. 
Very little work appears to have been done on the dimensions, 
numbers and distribution of the stomata of grasses, and none, so 
far as I have been able to discover, on those of Bromus or on the 
seedlings of any genus. 
It has been known since the time of Weiss 1 and even before 
him, that the stomata of grasses are arranged in longitudinal rows 
flanking the ridges, and that in some genera or species they occur 
only on one side, usually the upper, and in others on both sides of 
the leaf. In all the Bromes I have examined 2 they occur on both 
sides, and are more numerous on the upper than on the lower face 
of the lamina. Owing to their peculiar distribution on the flanks 
of the ridges it is not always easy to observe and count the 
stomata by direct observation of the faces of the leaf, since many 
of the stomata are seen obliquely from the side or are even more 
or less withdrawn from observation (unless great care is exercised) 
owing to the contractile action of the motor cells, which causes 
them to sink into the flanks of the narrow furrows as the leaf rolls 
inwards. 
It seemed that, since infection occurs by means of the stomata, 
an accurate knowledge of their sizes and numbers was, then, an 
imperative necessity, in order to test whether any relation existed 
between these factors and the susceptibility to infection of the 
species. 
Consequently, I undertook the very laborious task of count- 
ing the stomata per square unit of surface of the leaf, both above 
1 Pringsli. Jahrb. B. iv. 
2 With one exception, a species which requires further investigation in other 
respects. 
