THE INTERNAL STRUCTURE OP THE LEPIDOPTEROUS PUPA. 77 



cells, and sends a protoplasmic process upward to form the scale. This 

 is well exhibited in PI. i., fig. 6. The scale at this stage is a minute 

 flattened chitinous bag, filled with protoplasm, and, whilst the 

 scales remain full of protoplasm, they appear as transparent as 

 glass, but when the protoplasm shrinks out of them they become 

 whitish. 



The hypodermal cells, although no longer separated by well-defined 

 cell walls, are still well marked out by the peculiar arrangement of the 

 finely granular contents of the cells. The hypodermis, too, has now 

 begun to secrete the chitinous cuticula of the wing membrane ; but it 

 is as yet very thin, becoming much thicker as the wings develop. 

 Each of the hypodermal cells, at this stage, gives rise to a new 

 prolongation (jbr. h'drm.) from its base, which, piercing the grund- 

 membran below it, traverses the lumen of the wing, pierces the grand - 

 membran of the opposite side, and finally unites with the cuticula of 

 the opposite surface of the wing (fig. 8, fbr. h'drm.), and thus every 

 hypodermal cell becomes converted into a long, thin fibre, stretching 

 from the upper to the lower surface of the wing, and the similarity of 

 their appearance to muscular fibres suggests that they may be con- 

 tractile, although Mayer believes that they, in time, become tendinous 

 cords, serving to hold the opposite membranes of the wing together 

 during its great expansion, directly after the exclusion from the chry- 

 salis. Schaffer thought that these fibres were merely the original 

 protoplasmic prolongations (such as are shown in PI. i., fig. 5, pre.) 

 fused together, but he was quite unaware of the absorption of these 

 prolongations, and the subsequent development of the fibres uniting 

 the opposite membranes forming the wing. 



In the scale itself, the protoplasm which fills the primitive scale 

 undergoes contraction, and becomes coarsely granular, and gradually 

 withdraws from the scale, leaving little chitinous pillars (elm., PI. i., 

 fig. 11) which bind the upper and lower surfaces of the scale together. 

 Mayer says that the protoplasm is entirely withdrawn, and that the 

 scales then become merely little flattened, hollow, chitinous sacs, con- 

 taining only air. Chapman, however, dissents entirely from this 

 view, and states that no air enters any of the scales until the pigment 

 is fully elaborated. We have ourselves noted that the scales, at this 

 so-called " white " stage, are very different in appearance from the 

 opaque whiteness which characterises actually white scales filled with 

 air, at the final stage of development. It appears that the scales, at 

 this stage are not filled with air, but are filled Avith a clear secretion 

 from the haunolymph, containing all the necessary materials for going 

 through the chemical changes which result in pigmentation. The 

 striations of the scale may be observed, at this stage, to be due to a 

 series of parallel longitudinal ridges on the upper surface (PI. i., figs. 

 10-11), the under surface of the scale is provided with but few, and 

 these ill-developed, ridges. 



The next stage of development shows the wings of a yellow-ochre 

 colour, for the secretion from the ha?molymph, which fills the scales, 

 begins to undergo the chemical changes which result in pigmen- 

 tation. The formative cells now show great change, and the forma- 

 tion of the scales being completed, and there being no further use for 

 these cells in the economy of the insect, they undergo degeneration. 

 At this stage, certain scales (? androconia) situated either upon the 



