364 
SEXUAL SELECTION. 
Part II. 
invariably females. This is probably a case of reversion, 
lor in the true Libellulse, when the sexes differ in 
colour, the females are always orange or yellow, so 
that supposing Agrion to be descended from some pri- 
mordial form having the characteristic sexual colours 
of the typical Libelluke, it would not be surprising that 
a tendency to vary in this manner should occur in the 
females alone. 
Although many dragon-flies are such large, powerful, 
and fierce insects, the males have not been observed 
by Mr. MacLachlan to fight together, except, as he 
believes, in the case of some of the smaller species of 
Agrion. In another very distinct group in this Order, 
namely in the Termites or white ants, both sexes at 
the time of swarming may be seen running about, “the 
“ male after the female, sometimes two chasing one 
“ female, and contending with great eagerness who shall 
“ win the prize.” 52 
Order, Hymenojotera .— That inimitable observer, M. 
Fabre,’ J in describing the habits of Cerceris, a wasp- 
like insect, remarks that “ fights frequently ensue 
between the males for the possession of some parti- 
‘ cl *lar female, who sits an apparently unconcerned 
“ beholder of the struggle for supremacy, and when the 
victory is decided, quietly flies away in company 
“ with the conqueror.” Westwood 54 says that the 
males of one of the saw-flies (Tenth rcdiiuc) “ have beeu 
“ fo und fighting together, with their mandibles locked.” 
As M. I’abi'e speaks of the males of Cerceris striving 
to obtain a particular female, it may be well to bear in 
Kirby and Spence, ‘ Introduct. to F.utomology,’ vol. ii. 1818, p. 35. 
,J See an interesting article, “ Tire Writings of Fubre,” in 1 Nat. Hist. 
Keview,’ April, 1882, p. 122. 
‘Journal of Proc. of Entomolog. Soc.’ Sept. 7th, 18(13, p. 109. 
