218 
[October 
the “posterior” series of subobsolete dark spots which I have since 
observed in several species of Heteerina seems to indicate that there is 
normally also a fourth dark vitta there—the posterior one. Yet in 
all these groups there are scarcely any structural specific characters 
but those which are drawn from the reproductive organs. Now what 
possible advantage can it be to a S Hetaerina or to a S Gromphus or to 
a S Agrion, to have the colorational stripes on any part of its body 
arranged a little differently, or a little wider or narrower, or one or 
more of them confluent, abbreviated, interrupted, subobsolete or obso¬ 
lete ? On the principles of Sexual Selection, we can readily and natu¬ 
rally account for % % being often more highly colored or more highly 
ornamented than 9 9 5 but it is unconceivable to me that such minute 
differences as those above referred to could be appreciated by 9 9 . 
Neither can I understand how such minute differences as these could 
be of any use to the whole species, both % % and 9 9 , so as to be seized 
on, not by Sexual Selection, but by Natural Selection. Darwin has 
felicitously conjectured, with regard more especially to Yertebrates and 
Plants, that “ the most frequent cause of variability may be attributed 
to S 9 reproductive elements having been [functionally] affected prior 
to the act of conception.” (Grig. Spec. p. 15.) The affections above 
spoken of are structural, not functional; but it is not improbable that 
they may similarly give rise to a considerable amount of variation in 
coloration and perhaps occasionally even in structure. Prof. Owen has 
observed that “ the generative organs, being those which are most re¬ 
motely related to the habits and food of an animal, have been always 
regarded by him as affording very clear indications of its true affini¬ 
ties.” (Quoted 0?’. Spec. p. 361.) It is remarkable that in Odonata, 
although the form of the % reproductive organs differs almost in every 
species, yet that there is a certain family likeness throughout many, per¬ 
haps most, of the groups. No Neuropterist, for instance, could possibly 
mistake the % abdominal appendages of any species of Agrion for those 
of a Lestes, or those of a Lestes for those of a Hetaerina. I select the 
abov-e examples because here the classification is not, as in the subge¬ 
nera of the genus Gomphus, based upon the structure of the % abdo¬ 
minal appendages; and it cannot therefore be contended that I am 
arguing in a circle and saying that certain species belong to a separate 
group because their % abdominal appendages have a general resem- 
