78 BRITISH LEPIDOPTERA. 



nervures or near the outer edges of the wing, have a leucocyte enter 

 them ; the pigmented scales contain no leucocyte. 



The insertion of the scale into the wing membrane is brought about 

 by the filling of the narrow cylindrical stalk of the scale into a minute 

 close-fitting socket, which perforates the wing membrane (PL i., fig. 

 10). It is not set into a tube, as Landois supposed," nor did Mayer 

 discover anything resembling the " Schuppenbalg " described by 

 Spuler, for the insertion of the scales. 



The transverse folding of the wing membrane is very sharp, and 

 the latter is, in fact, thrown into a very regular series of closely com- 

 pressed folds (eta. al., PI. i., fig. 10), a single scale being inserted on 

 the crest of each fold. When the imago emerges from the chrysalis, 

 the hamiolymph (blood), within the wings, is under considerable 

 pressure, and this would tend to enlarge the wing into an inflated bag ; 

 but the hypodermal fibres (PL L, fig. 8), before described, hold the 

 upper and lower walls of the wings together, and the bag becomes a 

 flat, instead of an inflated one. In A. archippus, Mayer points out 

 that " the area of the wing of the imago is 8*6 times that of the pupa. 

 The wing of the newly-formed pupa has about sixty times the area of 

 the wing in the mature larva, so that it is evident that, in passing from 

 the larval state to maturity, the area of the wings increases more than 

 five hundred times." 



We have already seen that when the scale first appears it is only a 

 small protoplasmic cell, which very soon increases in size, and flattens 

 out, and finally assumes the outward shape of the mature scale. A 

 layer of chitin is then secreted over its entire outer surface, so that the 

 scale becomes a thin, flat chitinous bag, filled with protoplasm, the 

 chitin upon the upper surface of the scale being striated, the lower 

 surface smooth. Many scales have two sets of striae — a well-deve- 

 loped longitudinal set, and a finer transverse set. These striations 

 diffract the light, and give rise to the iridescent colours observed on 

 the wings of many Lepidoptera. 



So long as the scales remain filled with protoplasm, they are quite 

 transparent, but the protoplasm afterwards becomes coarsely granular, 

 and appears to give place to a secretion from the haemolymph, Avhich 

 contains the necessary material for the elaboration of the pigment, the 

 white coloration being different from the opaque whiteness notice- 

 able in air-filled scales. Mayer says that "the scales which are 

 destined to be white upon the mature wing are now completely formed, 

 and undergo no further changes, hence, ontogenetically speaking, the 

 white spots (? scales, J. W. T.) upon the wing are the oldest of all." 

 Here we would offer two notes of criticism : (1) That the white scales of 

 an insect's wing are of two entirely different classes. Some white scales 

 are quite richly pigmented (see Entom. Rec, vol. vi., pp. 35 et seq., 109 et 

 seq., 204 et seq.), e.g., the white areas of the wings of Melanargia 

 galathea, the white spots on the wings of Syrichthus malvae, and 

 numerous other examples, which have been already fully discussed." 

 (2) That the whole of the scales (pigmented and unpigmented) are 

 supplied with air, only in the final stage of their development. 



The secretion from the haamolymph ( = the " pigment factor " of 



* " The genetic sequence of insect colours." — British Noctnae and their Varieties, 

 vol. ii., pp. i — xviii. 



