226 
The only remaining element, it seems, with which to correlate these differ- 
ences is the season in which the broods were hatched. The characters of 
broods vary with the varying conditions of the years. Equally great changes are 
perhaps not uncommonly produced artificially in the laboratory by subjecting 
organisms to changed conditions during their ontogeny, so that there is nothing 
unusual in the changes occurring in these broods. But it is interesting to know 
that in nature within areas of such comparative uniformity as one of these small 
lakes the most uniform environment from year to year possible in this latitude 
there is a sufficient fluctuation in the seasonal conditions to produce these meas- 
urable differences found in these different broods. 
V. ARE THE VARIATIONS IN ONE FIN CORRELATED WITH THE VARIATIONS 
IN THE OTHERS? 
Most of the specimens from both lakes were separated into different groups 
on the basis (1) of the number of rays in the anal fin, and (2) of the number of 
spines in the dorsal fin. In the former grouping, by finding the average number 
of elements in the dorsal fins, both separately and combined, for each group, the 
correlation of the variations in the anal fin with that of the dorsal separate and 
combined, can be determined. Similarly in the latter grouping (2) the correla- 
tion between the two dorsals can be determined. 
In the way of an illustration, the data for the broods of ’95° and ’96° from 
Tippecanoe Lake are given in Table XV. The number of specimens in each 
group occurs in the first column. 
It will be observed that in both broods there is a definite correlation between 
the anal fin and dorsal fins, separately and combined. 
When the anal rays increase in number, the dorsal spines, dorsal rays, or both 
combined, also increase. There are several exceptions to this law, noteworthy in 
the dorsal spines of the twelve anal-rayed group and the dorsal rays in the eleven 
anal-rayed group of the brood of ’96°. 
In all cases the increase in the dorsal is considerably smaller than the increase 
in the anal. The correlation is stronger for the two dorsals combined than for 
them separately, and of the latter, it is the stronger for the dorsal rays. The 
ratio of increase in the dorsal to the increase in the anal is approximately twelve 
to two, five to two and four to two in the spinous dorsal, soft dorsal and both com- 
bined, respectively. 
