(*3eteaer 
the same shape but taper slightly, the distal end being the 
smaller; the ordinary scales around are not set up to this 
brand, but are rather overlaid by the androconia whilst the 
“mark” is crossed by two veins which modify both the colour 
(brown instead of black) and the size of the scales clothing 
them. Though the apices of the androconia are black this 
colour is only at the apices, the lower parts being pale brown; 
this however is by no means of infrequent occurrence. 
A discussion on the scales of the Ruralidae could not be 
closed without some reference to the unique and primitive 
though extraordinarily specialised genus Liphyra, with the 
delicate white fugitive scales attached to its wings on emer- 
gence from the pupa, and with the yet more peculiar body 
scales all of which probably disappear after the first flight or 
two; both are quite evidently protective developments. 
The long white scales of the wings are much the shape of a 
hly of the valley leaf with its end truncated and scalloped; 
in size they vary considerably, whilst the scalloping of the 
apex varies yet more; some have two scallops, others go up 
to four or five; the point of attachment is also peculiar, there 
being a twin attachment excavated between the two points 
which are almost at right angles to the stem, this being no 
doubt necessary to allow it to be easily freed from the ordinary 
scales; this is the usual method, but occasional scales are 
present with but one; they are all ribbed, the ribbing not 
being particularly fine. Under direct light the colour is pure 
white, but under transmitted light grey, more or less trans- 
parent. The body scales are however much more unique 
than these just described; the abdomen of the imago looks 
to be densely covered with silky fluff, the word fluff is the 
only word that adequately describes it, this however under 
the microscope is found to consist of scales of extraordinary 
size and structure; they are exactly like the egg case of the 
skate, but instead of having a long filament at each corner, 
they have one from each end in the centre, that at the one 
end being very long indeed, frequently much longer than the 
rest of the scale, whilst that at the other end, probably the 
end of attachment, is by no means as long and not infrequently 
has two points of adherence (7.e. Y shaped), the stem of 
