Very little is positively known as to the development of the colour in the wing-scales. It has been pointed out by Hopkins[199] that in some cases the colours are of the nature of urates; that is, of excretory matter of the kind that usually passes from the body by direct channels, and in the case of Lepidoptera, by the Malpighian tubes. Miss Newbigin suggests that the organic pigments used in scale-coloration will be found to be of two kinds, urates and melanins, the urates being derivatives from nitrogenous, the melanins from carbonaceous, matters.[200] Marchal, who has devoted a great deal of attention to the study of the Malpighian tubes, informs us that the subdermal pigments of caterpillars are frequently in large part deposits of urates, and he is of opinion that, the function of the Malpighian tubes being arrested at certain periods of the metamorphosis, elimination of the matter they separate when functionally active then takes place in a variety of other ways.[201] A similar condition as to the melanin-pigments and the respiratory functions appears also {335}probable. The scales when first formed are pallid, and the physiology of their pigmentation is not fully ascertained; it is, however, known that when the scales are pallid the hypodermis is either pigmented or in close contact with pigmentary matter, and that as the scales become coloured this pigmentation of the hypodermis diminishes; so that it is clear that the colour of the scales is obtained from matter in the interior of the developing wing, and probably by the agency of the hypodermis. The pattern on the wings of Lepidoptera is formed before the emergence from the pupa. In the Tortoiseshell butterfly, according to Schäffer, it commences to appear about the ninth day of the pupal life, and the pattern is completed about the eleventh or twelfth day. He also states that the process varies in its rapidity, and this, he thinks, may depend on the previous condition of the larva. According to Buckell the pupa of _Nemeobius lucina_ is sufficiently transparent to allow the development of the colour of the imago to be watched. He says that the coloration occurred first in front; that its entire production occupied less than twenty-four hours, and only commenced about forty-eight hours before the imago emerged.[202] When the butterfly leaves the pupal skin the wings are soft, crumpled sacs, of comparatively small size, but, as everyone knows, they rapidly expand and become rigid; the physiology of this process is apparently still unknown.