No. 409.] CASES OF SALTATORY VARIATION. 37 
The length of the normal barbelis indicated in the table by 
heavy figures. 
Summarizing these data, we get the following: 
Number of cases with two barbels: left, 3; right, 2. 
Number of cases with one barbel : left, 6 ; right, 7. 
Number of cases with a Y or T shaped barbel: left, 3 ; right, o. 
Number of cases with no supplemental barbel : left, o ; right, 1. 
Number of cases with one on the left and none on the right, 1. 
Number of cases with one barbel on each side, 4. 
Number of cases with two barbels on each side, 1. 
Number of cases with two barbels on the left and one on the 
right, 2. 
Number of cases with one barbel on the left and two on the 
right, I. 
Total length of all supplemental barbels, not counting minute 
lobes, XV : left, 374 ; right, 244. 
Average variation, B left, 4.166 ; right, 2.722. 
While it is difficult from the variety of elements (number, 
length, shape) that enter into each variation to make a direct 
estimate of the degree of variability of each side and the degree 
of correlation of the variation of the two sides, the number of 
individuals is so small that an approximation can be arrived at 
by a glance at the data. 
The left side is much more variable than the right. 
The presence of barbels on the two sides is an indication of 
bilateral correlation. This is reinforced by the fact that in 
five out of nine cases the number of barbels on the two sides 
is the same. It would seem that the left side is leading 
in the addition of barbels, in spite of the fact that in one 
specimen no supplementary barbel is found on this side. 
Admitting, in the absence of evidence to the contrary and 
strong probability in its favor, that the specimens belong to 
one brood, or, at the very least, that the specimens occupying 
the narrow limit of environment in which they are found are 
