( ee yo 
I am well aware that the above sketch is extremely im- 
perfect, and that the actual facts are enormously more com- 
plex than I have put them, and that there are other things 
that ought to be taken into consideration: but I am also 
aware that there are conditions that tend to intensify the 
action of the factors I have mentioned. 
M. Fabre has devoted one chapter of his delightful work 
on the habits of Hymenoptera to a special argument against 
transformism, and gives it the rather witty heading, “a sting 
for transformism.” I have not been able to gather very dis- 
tinctly from his writings what he means by transformism ; 
but in the chapter I am alluding to his argument leaves no 
doubt at all that he is referring to the theory of descent of 
several existing species from a common ancestor. I have 
myself elsewhere stated that I am inclined to think that 
theory will prove to be to a great extent erroneous, and have 
given reasons for my opinion ; I should therefore have gladly 
welcomed M. Fabre’s support, if I found it possible to accept 
his argument as good, but, I regret to say, I fear it is likely 
to prove invalid: it amounts briefly to this, that if several 
species of Scolia that now feed their young each with a 
different and distinct kind of insect-food are descended from 
a common ancestor, then that ancestor must have been in 
the habit of feeding its young, not with some particular 
species of insect-food, but with a variety of kinds; and, if 
this was the habit of the original Scolia, M. Fabre urges that 
its offspring could never have abandoned this advantageous 
course of living, which he calls ‘‘ régime varié,” for the habit 
of living on a single species, or ‘‘ meals of a single dish,” as 
he styles it. The basis of this argument is that régime varié 
is a great advantage to the insect; if it is not, then 
M. Fabre’s argument fails completely,—the sting proves to 
be without point or venom. He cites instances to show that 
régime varié is an advantage, but every one of them is taken 
from the Vertebrate class. Now, it may be highly probable 
that régime varié is an advantage in the case of Vertebrata, 
but is it so in the case of insects? I fear we must say 
probably it is not: in the first place, it is comparatively rare 
among insects, one of their peculiarities being their extreme 
