22 THE AMERICAN NATURALIST. [VorL. XXXV. 
posterior borders of the first costal plates, and between the 
posterior edge of the carapace and the transverse line marked 
by the anterior borders of the next to the last pair of costal 
plates. The rest of the carapace is made up of four bony seg- 
ments instead of the five which are typical of normal individ- 
uals (Fig. 1). In this abnormal specimen, therefore, a whole 
bony segment (a neural plate, a pair of costals, and a pair of 
marginals) is absent, and the position of the deficiency is some- 
where between the posterior edge of the first pair of costals 
and the anterior edge of the next to the last pair of costals, 
c, in the body of the carapace. 
Such a suppression might be expected to be accompanied by 
a shortening of the carapace, and, as a matter of fact, there is 
some evidence that this is so. The length of the carapace 
under consideration is 1 5.7 cm., its breadth 12.1 cm.; the 
length is, therefore, 1.298— times the breadth. In ten normal 
males taken at random the average length of their carapaces 
was 16.65 cm., and the average breadth 12.68 cm. ; the average 
length was therefore r.313 + times the average breadth. It 
thus appears that the abnormal specimen is somewhat shorter 
than the average normal specimen, both absolutely and rela- 
tively to its length, — a condition to be expected if the suppres- 
Sion of a segment is assumed. 
Although the second specimen shows abnormal conditions 
in both scutes and bony plates, the fact that the two abnormal 
regions do not overlap even in part (for the most posterior 
position assignable to the bony-plate abnormality is still ante- 
rior to the most anterior position assumable for the scute 
abnormality) might well lead to the inference that these two 
irregularities were in no true sense correlated. Such a con- 
clusion, however, is probably incorrect. The fact that neither 
of the two abnormalities interferes with bilateral symmetry 
at least suggests something more than an accidental relation, 
though it in no wise meets the objection that the abnormalities 
are not superimposed. This objection, however, is not as 
serious as at first sight it seems to be. Harrison (998) some 
time ago showed by a series of ingenious experiments on 
young tadpoles that in the growth of the posterior parts of 
