Jan.— Feb. 1SS3.] 



PSYCHE. 



11 



Glycerin produces nearly the same 

 effects in these scales as water does ; its 

 action is, however, slower and gives 

 more opportunity to examine the scales 

 during the process of saturation. Light 

 transmitted through scales that are 

 treated with glycerin is nearly color- 

 less, and reflected light from scales 

 thoroughly penetrated with glycerin is, 

 for the most part, colorless ; sometimes, 

 however, it is greenish in spots, and 

 these spots are usually where a tinge of 

 red remains, when viewed by trans- 

 mitted light. 



In absolute alcohol the scales show, 

 by transmitted light, a more reddish 

 tinge tlian they show in water : in 

 chloroform the tinge is more purplish 

 than in water, altho the pin'plc is very 

 pale ; in turpentin and in oil of cloves 

 the scales are transparent and colorless. 

 Scales moistened with any of these 

 reagents and put over a dark or black 

 surface are light metallic green ; if dry 

 scales are put over dark surfaces they 

 are light metallic blue. Redried from 

 water, alcohol or chloroform these scales 

 regain their original colors, showing 

 that it is a coloration due to the struc- 

 ture of the scales and not due to any 

 pigment in them. In further proof of 

 this I exposed the scales to dry chlorin 

 gas and the color remained unaltered. 

 Solutions of chlorin or of hypochlorites 

 fail to destroy the coloration, for after 

 they are washed out and the scales 

 again dried the colors reappear as bright 

 as at first. 



Drv scales heated slightly over flame 

 suddenly lose all their metallic colora- 

 tion, anti, while retaining their form. 



become brownish grey by transmitted 

 or reflected light ; they are apparently 

 charred sufficiently to lose tlieir color 

 without having their structure greatly 

 altered. 



In structure these scales are readily 

 seen to be little flattened sacs ; wherever 

 they are injured, especially if they are 

 broken off near the base, the edges of 

 the upper and lower sides appear dis- 

 tinctly. The inner structure of these 

 scales is not easily discoverable. They 

 appear to be filled with a very delicate 

 network (fig. 3, </), which is always 

 reddish after the yellow has left the 

 scales. If glycerin is used as a reagent 

 in treating the scales their reddish net- 

 work, even in scales originally yellow, 

 remains long after the yellow has dis- 

 appeared. The yellow occupies the 

 interspaces of the network. The net- 

 work itself, under high jjovvers, appears 

 as if it were caused by retiform designs 

 in reddish oil between the layers of the 

 scales, but the fiict of the reappearance 

 of the color after treating the scales with 

 solvents for oil, e. g., chloroform, 

 shciws that it is not oil. From the ap- 

 pearance of scales charred to different 

 extents I am inclined to think this net- 

 work to be formed by thickenings of 

 the chitin walls of the scales themselves. 

 If this be the case, the thickenings pro- 

 ject only inward from these walls, and 

 are found in corresponding figures upon 

 both upper and under walls of the 

 scales. The colors red and yellow by 

 transmitted light may exist where only 

 the upper or under half of the scale is 

 present. 



{To he continued.') 



