( chexxii 
each being of moderate length; these scales also vary much 
in size and in the length of the filaments; in the smaller ones 
the likeness to the “‘ mermaid’s purse’ diminishes, they are 
rather coarsely striated, the striations being by no means 
sufficiently uniform to be called ribbing. In colour they are 
golden brown under direct light, and are much the same hue 
by transmitted light. When it is remembered that the whole 
abdomen is covered with these scales, their number must 
run into very many thousands; their size and arrangement 
must make them almost perfect for their purpose, so that 
when ants attack the newly emerged butterfly the long 
filaments at each end must be most admirably adapted for 
curling round and adhering tenaciously to their antennae, 
and thus ensuring a hasty retreat. 
Inasmuch as all the metallic greens of the Ruralidae, when 
seen under transmitted light, showed that they had no green 
pigment but that their basis was mauve or pink to red, it 
became necessary to investigate the allied colours in other 
genera. I therefore first looked to the great family Papilio 
and made several preparations, among others, of Papilio 
blumet. This is a very brilliant bluish-green species. I 
found the general pattern of the scales to be very diverse from 
those of the Ruralidae; the shape is pyriform, the scales being 
very coarsely reticulated, there being only ten rows of reticula- 
tions on these large scales (in a few very broad scales I counted 
as many as thirteen rows); the ribbing and the transverse 
ribbing are very deep and very irregular, being roughly 
quadrangular, but the quadrangles are of very varied sizes 
and they appear to contain one globule of (I suppose) pigment 
of a magenta colour (under transmitted bright light); in 
examining this, I reflected a very intense white light through 
the scales, the result being that the lines of the reticulations 
were finely buff colour whilst the contents were pinkish 
magenta; absolute sunlight, reflected through, left the same 
colours only intensely clarified; an ordinary but clear light 
produced dull ochreous lines and dull magenta colouring 
inside the lines; it will thus be seen that the general colouring ~ 
was the same in all lights, it being simply a question of degree 
of brilliancy. Here again there was no bright bluish green 
