256 The American Naturalist. [March, 
Comstock’s essay, relating to two kinds of characters ; first, char- 
acters indicating dichotomous divisions of lines of descent; 
second, characters indicating degrees of divergence from a prim- 
itive type. It is evident that either one of these kinds of char- 
acters may be subordinate to the other, although, at first glance, 
perhaps, it might seem that characters indicating differences in 
kind of specialization, or dichotomous divisions of descent 
lines, must always be superior to characters indicating differ- 
ences in degree of the same kind of specialization. For ex- 
ample, given a dichotomous branching according to characters 
of the first kind, the forms along either line will be ranged 
according to characters of the second kind, i. e., differences in 
degree of specialization along that line. But,ina larger view, 
there is, in the development of any considerable group of or- 
ganism, as, for example, the class Insecta, a general tendency 
of specialization along some pretty distinct main line, or more 
or less nearly parallel lines. For example, in the Insecta may 
be adduced the development of the flight function accompan- 
ied, in the Lepidoptera, by a cephalization of flight indicated 
by the specialization of the front wings and a reduction of the 
hind wings and accompanied also by the specialization of the 
thorax in the manner pointed out in this paper. Subordinate 
to any general tendency, such as the development of the flight 
function, there will appear characters indicating dichotomous 
_ divisions of lines of descent, the methods of advance along the 
line of the general tendency differing in two branches of the 
_ group. An example of this is afforded by the Odonata and 
the Diptera; in one group the specialization of fore and hind- 
wings has followed the same lines, in the other the specializa- 
tion has resulted in the loss of the hindwings. In both in- 
stances a fine development of the flight function has been 
reached. 
Of the kind of characters indicating in a general and large 
way degrees of divergence from a primitive type, these char- 
acters drawn from the wing-clothing and thoracic structures 
may be looked on. 
