116 
ZOOLOGY: W. M. WHEELER 
myrmica chamberlini has lost its wings, but apparently so recently that 
it still retains the typical male structure of the head, antennae and thorax, 
even to the development of the Mayrian furrows. In several other 
genera (Formicoxenus, Cardiocondyla, Ponera), however, the males 
have acquired the same structure of the head and thorax as the worker, 
so that they can be distinguished only by their genitalia and the num- 
ber of their antennal joints. In one species (Aner gates atratulus) the 
apterous male degenerates still further into an almost pupoidal condition. 
The facts briefly presented in the preceding paragraphs seem to me 
to have an important bearing on the question of continuous variation 
versus mutation in the production of organic forms. In most species 
of ants the constant and striking structural differences between the 
different castes would, at first sight, suggest that such forms as the 
apterous females, apterous males, soldiers and workers, must have 
arisen as so many saltatory variations, or mutants and that they sur- 
vived and secured representation in the germ-plasm, because they 
happened to fulfill specialized and useful functions in the life of the 
colony. I believe, however, that this view of the castes, at least so 
far as their origin is concerned, cannot be maintained, because all the 
available evidence points to their being merely the surviving extremes 
of graduated and continuous series of forms, the annectant members 
of which have suffered phylogenetic suppression or extinction. This 
is most clearly seen in the case of the soldier and worker. Only within 
comparatively recent time, i.e., probably since the middle Tertiary, 
has the originally monomorphic worker caste become polymorphic 
in certain genera (Camponotus, Atta, Fheidologeton, some species of 
Pheidole), i.e., developed a complete series of workers ranging from 
huge-headed major workers or soldiers (macrergates, dinergates) 
through intermediates of various sizes (desmergates) to small workers 
(micrergates) . There is much evidence to show that in some genera 
(e.g., Pheidole, Oligomyrmex, etc.) all the forms in this series, except 
the dinergates and micrergates, have been suppressed, so that a marked 
dimorphism of the worker caste, simulating an origin of one or both 
of the forms by mutation, has been produced. In other genera {Care- 
bara, many Solenopsis) the soldier form has also been suppressed, so 
that the worker caste has again become monomorphic through the sur- 
vival of nothing but the smallest forms (micrergates) of the originally 
graduated series. Finally, in certain parasitic ants, {Aner gates, Aner- 
gatides, Epcecus, Wheeleriella, etc.) the last traces of the worker caste 
have vanished, just as in several Australian genera (Leptomyrmex, 
Rhytido ponera, Diacammay^ and the South African Ocymyrmex^^ the 
