Page:Popular Science Monthly Volume 68.djvu/303

Rh of the female sex in ants—including under this term both the fertile, or queen phase, and the usually sterile, or worker phase—reaches its clearest expression in the extraordinary range of intraspecific polymorphism. In certain species, for example in the African driver ants (Dorylii) and American ants of visitation (Ecitonii), the structural differences between the workers of the smallest caste and the huge queen of the same species are enormous and represent an amplitude of variability in the female sex unequalled in any other organisms. Male ants, on the contrary, exhibit so little variability that it is often difficult, or even impossible, to distinguish the genera of single specimens of this sex. These facts have an important bearing on the views of authors like Brooks and Geddes and Thomson who assume that male animals are more variable than females, and of those authors who have transplanted this hypothesis to the fields of sociology and anthropology. All of these writers maintain a discreet and significant silence on the subject of the social insects. Equally astonishing, however, is the attitude of the biometricians, who, priding themselves on the accuracy of their methods and repudiating mere observation and speculation, proceed to an elaborate measurement of the wings of honey-bees and ants for the purpose of ascertaining whether males are more variable than females, when a glance at the personnel of a few ant and termite colonies would convince the most skeptical that there can be no such correlation between sex and variability as that assumed by the above-mentioned authors. If it is clear that the males of many of the higher animals are in certain characters more variable than the cospecific females, it is even clearer that the very opposite is true of the social hymenoptera, while in the termites, or white ants, both sexes seem to be alike variable and polymorphic.