112 THE BUTTERFLIES OF CEYLON 
in making this examination, and great experience in estima- 
ting the value of the characters observed, are necessary to 
form an opinion on the subject.”” I confess that I have had 
no previous experience, but the sketches in the plates 4, 5, 
and 6, are in nearly every case the result of the examination 
of a large number of specimens. In no case was the sketch 
made from a single specimen. I have sent sets of my slides 
to the Colombo Museum and to the Indian Agricultural ° 
Institute at Pusa, where any one interested can examine 
them. They include all the Ceylon Hesperiidae, with the 
exception of C. spilothyrus and G. albofasciata. When 
removed from the body and allowed to dry, the clasps often 
shrivel and curl up, thus entirely altering their outlines 
as seen under a microscope. I have, therefore, whenever 
possible, used perfectly fresh, undried specimens for my 
sketches. 
In cases where there are only slight differences between 
the prehensores of two forms, it is advisable to examine a 
large number of each, to ascertain if these differences are 
permanent, or only casual variations. 
Unfortunately my material from India has been, so far, 
limited to about 500 specimens, and in only a few cases have I 
had a good series of any one form. My experience has been 
that, in some genera, notably Telicota and Halpe, the species 
. show a strong tendency to split up into local races and I am 
of opinion that, in these cases, the prehensores will be found 
a more constant and reliable guide than external colouration. 
I have dissected very large numbers of some Ceylon species 
and have always found that the variation of the prehensores 
is very slight. 
A question which arises is whether differences in colour 
caused by climatic influences are accompanied by changes in 
the prehensores. Recent investigation seems to prove that 
they are. Fruhstorfer found seasonal variation in the clasps of 
Nacaduba pavana, and other instances are known. Indian 
writers treat Caprona saraya as a seasonal form of C. ranson- 
nett, though Elwes and Edwards point out that their clasps 
