578 Wm. S. Marshall, 



less covered with short setae except over that part where the wing 

 rudiment appears, here the chitin is smooth. 



Newly hatched larvae and those only a few days old woiild average 

 1.5 mm in length. In the material collected there were no specimens 

 between these veiy young larvae in which no trace of the wing rndiment 

 could be found in the hypodermis and specimens that had attained 

 a length of 3.5 mm. In these latter the area of the wing rudiment 

 could be distinguished both by the darkened cuticula and by the under- 

 lying hypodermis which, at this stage, differed from that covering 

 the rest of the thorax. 



In larvae 3.5 mm long the hypodermal layer of the second and 

 third thoracic segments was of a nearly uniform thickness throughout 

 and generally a little thicker than the transverse diametre of any of 

 its nuclei. Within the area of the wing rudiment the nuclei of the hy- 

 podermal cells were more abundant than within an equal area taken 

 from any other part of the thorax the nuclei being so crowded together 

 that adjacent ones often appeared to touch (Fig. 2). In the hypo- 

 dermis of the youngest larvae, just hatched, the elongated nuclei were 

 found to lie flat against the cuticula, in most of the larvae 3.5 mm 

 in length the nuclei were irregulär in outline with many of the more 

 elongated ones occupying a position in which the longitudinal axis 

 was radial to the cuticular surface. No dividing nuclei were found, 

 in larvae of this length, in the wing rudiment and the absence of short 

 setae on the cuticula of this region was very noticeable. From the 

 base of the hypodermal cells within this area, especially those near 

 the margin, there were often thin, threadlike Strands of cytoplasm 

 stretching to other, adjacent, parts such as the fat body or the near 

 by hypodermal cells. 



The next larvae examined were 5 mm long ; a few of these showed 

 no change from what has just been described although, in nearly all 

 of them, the hypodermis in the area of the wing rudiment was thicker 

 than at other parts of the thorax. The nuclei of these hypodermal 

 cells showed a increase in size and nearly all of those within the area 

 of the wing rudiment now occupied such a position that the longitu- 

 dinal axis of each was radial, or nearly so, to the surface of the cuti- 

 cula (Fig. 3). Within this same area there were proportionally more 

 nuclei but mitotic figures were seldom seen. Nuclear division does 

 not appear to be rapid in the hypodermis but must proceed more 

 räpidly within the area of the wing rudiment than over other parts 

 of the thorax. While these few changes are taking place the nuclei 



