(, keivai 
like scales of the median area are quite different in several 
respects. They are many times as thick, very long, and 
waved, attached to the wing not by a short peduncle or stalk, 
but by a slight spreading of the base directly on to the 
membrane, the basal part of the scale is colourless, slightly 
swollen, soon becoming shortly constricted, at which part 
the brown pigment developes and the very fine ribbing begins ; 
beyond this it very rapidly thickens slightly and remains 
of even width until near the apex, when it shortly tapers to 
a blunt point. With these scales, especially on the veins 
is found another from which it may be possible the former 
were evolved; it is unusually long and narrow, but not hair- 
like, it very gradually tapers wider to about a fifth from the 
apex, when it rapidly reduces to the tip which is abruptly and 
almost squarely terminated; the attachment stalk is also 
specialised, being between the ordinary method and that 
obtaining in the hair-like brown scales; it is thicker, and gradu- 
ally tapers for a short distance to its juncture with the scale, 
whose base is simple, not ribbed, but again tapering up 
shortly to its juncture with the part where the usual ribbing 
arises, and from whence the rest of the scale is finely ribbed; 
these two joints (if I may use the term) are the nearest 
approach to hair structure that I have yet found, and the 
interesting part is that they do not occur in the hair-scales, 
but in a scale proper. The question that naturally arose 
was, will there be a modification of the usual wing scales 
that obtain under the brown hair-like scales? At the extreme 
base where the ordinary whitish hair scales still persist there 
was no modification, the scales presenting nothing worthy of 
note and having sharply serrated apices, but in the median 
area this was not so, the blue scales being of the simplest 
pattern; they are of a long oval shape, the apices being of 
two patterns, one quite evenly rounded, the other with a 
slight shoulder on each side, the apical edge itself however 
being quite evenly rounded off; these latter are very few in 
number. In the postmedian area the blue scales are much 
shorter but with similarly modified apices, so that one is 
led to wonder whether the brown hair-like scales were not 
at one time more largely spread over the wing. The “ blasen- 
