RECENT LITERATURE. 103 
for descriptive purposes must surely lay them aside when they look 
at the figures in this volume, where, in the three hundred odd figures, 
it would be impossible to find one that could be confused with any 
other. Mr. Pierce says he sees no difference between favicolor and 
pallens except size, yet if his figures can be relied on there is consider- 
able difference in some of the proportions and in the spinous 
armature of the cucullus. It is a case like this in which a photo- 
graph, even a poor and imperfect one, would much exceed in value 
any drawing. There are, however, two circumstances which make 
the appendages rather more valuable than any other structure, or 
even than a combination of other characters, for the decision of a 
question either of specific distinction or of generic or familiar 
alliance. One of these is that so long as two forms remain one 
species, that is, are syngamic, or breed together, directly or in- 
directly, these parts are kept identical in both. As soon as they 
become distinct species, divergence of the parts is liable to occur, 
perhaps rapidly. Only syngamy can keep these structures uniform 
throughout a group of forms. When syngamy ceases, they are free 
to vary, the only necessity being that they remain identical through- 
out each syngamic group. But their variation is not restrained by 
any such question as adherence to a food-plant, avoiding particular 
enemies, &c., as nearly all other characters used to define species are 
characters that are kept constant by natural selection. The other 
circumstance is that they are highly organised and for the most part 
hard, chitinised structures, so that they afford many details for 
observation and these details are embodied in forms less difficult to 
seize than many such items as colour, wing markings, «ce. 
It is now well known, Dr. Jordan having perhaps most clearly 
pointed the fact out, that geographical races have these appendages 
varying a little in each race. In some cases, when the segregation 
is great, these no doubt mark incipient species ; when the segregation 
is incomplete, as at the extremities of an extended habitat, there may 
be a difference that is prevented from becoming great by mediate 
syngamy. 
When there is immediate syngamy throughout a group of forms 
(unquestionable species) we may expect to find great uniformity 
throughout the group, but it is highly probable that in such a group 
variations may take place very rapidly, the whole group moving 
together. This seems a conclusion inevitable from the many instances 
we meet with in which the appendages are extremely different in 
closely allied species. 
Unquestionably the appendages, like everything else, are under the 
control of natural selection, but as to what the circumstances are 
that tend to govern the selection, we are certainly at present very 
much in the dark, so that if our argument seems to suggest that 
they have escaped its control, all that is postulated is that they are 
not controlled by any of those items of environment that have to be 
counted on by most other characters, and that as a result two 
newly established species may retain by force of similar environment 
a very similar facis and structure in everything but these structures, 
which are free to vary to any extent if only they vary together 
throughout each specific group. 
