324 Annals Entomological Society of America [Vol. VIII, 
the disappearance is not accomplished gradually, by successive 
attenuation or evanescence of the cocoon, but suddenly, so to 
speak, since larve that are similar and give rise to similar 
adults, make an absolutely perfect cocoon, which shows no 
signs of reduction, or make none at all. This may be cited as 
an example of the sudden changes that may supervene in the 
habits of an animal. It illustrates the conclusions embodied 
in a communication made by my brother, M. Armand Janet, 
to' the Leyden Congress. These conclusions, deduced from 
considerations of rational mechanics applied to the problem 
of the species regarded as a position of equilibrium, tend to 
show that the differences obtaining between a form and its 
derivatives are more probably produced by rather sudden 
saltations than by insensible and continuous variations.”’ 
In other words, Janet regards the disappearance of the 
cocoon in ants as due to what we now call ‘‘mutation,’’ and 
intimates that in certain species the process is still going on. 
I am willing to grant the partial truth of this contention, 
but I have recently found evidence. to prove that, far from 
being a recent loss, the disappearance of the cocoon, in certain 
genera of ants at least, had already taken place not less than 
two million years ago. Among the Dolichoderine of the 
Baltic amber, which is of Lower Oligocene Tertiary age, I find 
the pupe of certain species already lacking the cocoon, and 
this structure is likewise absent in a species belonging to the 
Camponotine genus Prenolepis (P. henschet Mayr) in the 
same formation (1914). Facts like these are calculated to 
impress us with the immense antiquity of many apparantly 
trivial structures and to convince us that we cannot dispense, 
even in these days of experimental biology, with a certain 
amount of ‘‘ phylogenetic speculation.”’ 
What has led to the disappearance of the cocoon in the 
Myrmicine, Doryline and Dolichoderine and in a few genera 
and sporadic species of other genera among the Camponotine? 
This question no doubt admits of several answers. Janet, as 
we have seen, believes that the cocoon became superfluous and 
disappeared because it was replaced by the fostering care 
and protection of the workers of the colony. But if this is the 
case, why do the most formidable of all ants, the Australian 
‘“‘bull-dogs’’ (Myrmecia), retain the cocoon and such feeble 
