MAYER: DEVELOPMENT OF WING SCALES. 219 
which comes in contact with the Grundmembran and fuses with it 
(Plate 4, Fig. 27). Then a bundle of these processes, elongating still 
more, breaks through the membrane, traverses the lumen of the 
wing, and fuses with the Grundmembran of the opposite wall (Fig. 12, 
fbr. Wdrm.). Indeed, the growth does not stop here, for the bundle of 
fibres pierces this second Grundmembran, and effects an attachment to 
the cuticula of the opposite surface of the wing (Figs. 11, 13, fbr. h’drm.). 
Very soon every hypodermis cell becomes converted into a long thin 
fibre stretching from the upper to the lower surface of the wing. When 
these fibres are stained, and cut in cross section, it is seen that the 
central core of the fibre remains almost colorless, while the peripheral 
portion, which is more highly refractive, stains deeply. As Dr. Mark 
suggested to me, the appearance is strikingly similar to that presented 
by the muscle fibres of many invertebrates, and it is therefore possible 
that these fibres may be contractile. It is, however, especially to be 
noticed, as he also remarked, that these fibres have never been observed 
to present the transversely striated appearance common to all the 
known muscles of insects. It is highly probable that they, in time, be- 
come tendonous cords, which serve to hold the two walls of the wing 
membrane close together during the great expansion of the wing which 
occurs upon emergence from the chrysalis. Schiffer (89, p. 645, Tafel 
XXX. Fig. 39) observed these fibres in a cross section of the wing of a 
well advanced pupa, a recently emerged insect, and also in a pupa 
immediately before emergence. But not having had material inter- 
mediate between this and the earliest stages in the formation of the 
scales, his notion of the manner in which these fibres are formed is 
apparently inaccurate ; for he seems to assume that they are merely the 
primitive protoplasmic processes of the hypodermis cells, such as are 
shown in Figure 5, pre. His idea is that the Grundmembran must 
become absorbed in some way, thus —as his Figure 39 would seem to 
indicate — allowing the protoplasmic processes (i.e. pre., Fig. 5) of op- 
posite surfaces of the wing to fuse together and form little pillars (each 
with a nucleus at ether end), which bind the two surfaces of the wing 
together. Since, as I have said, he was unacquainted with any of the 
stages of development between those corresponding with my Figures 8 
and 13, he failed to observe the gradual absorption of the protoplasmic 
processes (pre., Fig. 5), and the subsequent formation of the hypodermal 
fibres (fbr. h’drm., Fig. 12). 
But to return to the discussion of the stage represented in Figure 12. 
The protoplasm which once completely filled the scales has begun to 
