That the cylindrical form of the bee grub and caterpillar is the result of modification through descent is evident in the caterpillar-like form of the immature Caddis fly (Pl. 3, fig. 2). Here the fundamental characters of the larva are those of the Corydalus and Sialis and Panorpa, types of closely allied groups. The features that remind us of caterpillars are superadded, evidently the result of the peculiar tube-inhabiting habits of the young Caddis fly. In like manner the caterpillar-form is probably the result of the leaf-eating life of a primitive Leptiform larva. In like manner the soft-bodied maggot of the weevil is evidently the result of its living habitually in cavities in nuts and fruits. Did the soft, baggy female Stylops live exposed, like its allies in other families, to an out-of-doors life, its skin would inevitably become hard and chitinous. In these and multitudes of other cases the adaptation of the form of the insect to its mode of life is one of cause and effect, and not a bit less wonderful after we know what induced the change of form. Having endeavored to show that the caterpillar is a later production than the young, wingless cockroach, with which geological facts harmonize, we have next to account for the origin of a metamorphosis in insects. Here it is necessary to disabuse the reader's mind of the prevalent belief that the terms larva, pupa and imago are fixed and absolute. If we examine at a certain season the nest of a humble bee, we shall find the occupants in every stage of growth from the egg to the pupa, and even to the perfectly formed bee ready to break out of its larval cell. So slight are the differences between the different stages that it is difficult to say where the larval stage ends and the pupa begins, so also where the pupal state ends and the imago begins. The following figures (205-208) will show four of the most characteristic stages of growth, but it should be remembered that there are intermediate stages between. Now we have noticed similar stages in the growth of a moth, though a portion of them are concealed beneath the hard, dense chrysalis skin. The external differences between the larval and pupal states are fixed for a large part of the year in most butterflies and moths, though even in this respect there is every possible variation, some moths or butterflies passing through their transformations in a few weeks, others requiring several months, while still others take a year, the majority of the moths living under ground in the pupa state for eight or nine months. The stages of metamorphosis in the Diptera are no more suddenly acquired than in the bee or butterfly. In all these insects the rudiments of the wings, legs, and even of the ovipositor of the adult exist in the young larva. We have found somewhat similar intermediate stages in the metamorphoses of the beetles. The insects we have mentioned are those with a "complete