No. 401.] THE FRUITING OF THE BLUE FLAG. 379 
open for two days. Sometimes there is an interval of a day or 
two with no flower open. Some striking differences in time 
of flowering, due to situation, will be noticed under the next 
heading. 
I collected 214 stems from five different localities and stud- 
ied them to determine the proportionate fertility of the several 
capsules. Wholly infertile stems I disregarded altogether, and 
capsules developing any seed at all I counted fertile ; Table II 
summarizes the results of the count. Disregarding localities 
for the present, the total percentages of fertility are as follows : 
Flower Ja, 75 per cent; Z b, 63 per cent; I c, 28 per cent. 
u Z1 a, 66 per cent ; IT b, 43 per cent. 
*- Za, 68 per cent; LIT b, 54 per cent.! 
Clearly, the advantage is with the earlier flowers. 
But is there not further significance in these facts? In con- 
nection with the occasional occurrence of an extra flower in 
clusters / and // they seem to me to suggest that the flower 
cluster is in process of reduction by elimination of the later 
flowers. The observations of one season will not determine 
such a matter. It is sufficient to record the conditions of the 
present, that the future student may have the criteria necessary 
for determining it. 
VII. Tur RELATION oF HABITAT TO FERTILITY. 
My observations were all made on the eastern slope of the 
narrow Lake Forest moraine, which forms a part of the west 
shore of Lake Michigan, its wave-worn eastern edge rising 
abruptly some seventy feet from the water. Numerous sharp, 
post-glacial ravines, whose puny streams take their origin 
among the pools, marshes, and “ potholes” of the crest of the 
moraine and reach the lake level only at their mouths, give con- 
siderable variety to the otherwise uniform and scarcely percep- 
tible eastward slope. I found and studied the blue flag growing 
in eight different sorts of situations. ^ These will be referred to 
by numbers in the succeeding tables, as follows : 
! The higher percentages for cluster ZZZ are somewhat misleading, because 
this cluster is developed only under the most favorable conditions. — . 
