GO 
and, owing to persistent attacks, had to abandon a more lengthened 
stay. I have since been asked, “ Were all the butterflies worth the 
penalty?” Lying in one’s hammock with one’s temperature up to 
104 or 105°, and feeling as if it was immaterial whether the world 
came to an end, one would say no; but with a return to health one 
looks on the penalty with a light heart, and feels inclined to risk any¬ 
thing, and I should boldly answer yes to any such query now. No 
words, and mine least, can convey an adequate idea of what collecting 
in these forests is really like till one tries it. 1 can only say in con¬ 
clusion it has been a very real pleasure to write this paper. It has 
felt like living the time over again, and 1 hope I have transported you 
with me for the time being. 
THE IMPORTANCE OF CERTAIN LARVAL CHARACTERS AS A 
GUIDE IN THE CLASSIFICATION OF THE SPHINGIDS. 
(Head March 4th, 190*2, hy A. BACOT, F.E.S.) 
A few years since 1 wrote a short paper dealing with the possible 
relationship ol 7 Htnorpha ( I'.nilroinix ) rcrxicolora and the Sphingids 
(hut. Iiec. , vii., pp. 217-1246). Since then 1 have had the opportunity, 
thanks to the kindness of Dr. T. A. Chapman, Mr. J. W. Tutt, and 
many other friends, of examining a few more Sphingids and many 
other lepidopterous larva*, and my conclusions as expressed in the 
above paper have undergone considerable modification in consequence. 
Many of the points of resemblance that I there referred to I have 
since found to be common, with some modification, to other, and, in 
some instances, widely separated groups. So that, while regarding 
them as signs of a relationship, it is in a looser and broader sense 
than I had previously imagined. Indeed, so widely divergent are some 
of these groups from each other, that there can be, l take it, no 
question of near relationship, and we are forced to regard them as 
parallel developments to meet similar needs. Just as analogous 
developments have arisen independently in Marsupials and Placentals, 
so we find the same specialisation arising among groups of lepidop- 
tera having Hat eggs and those having upright ones, a difference that 
is possibly as important as that separating the placentals from the 
marsupial mammalia. 
To take a few instances that were referred to in my previous paper 
assigns of relationship between 7 Umorpha (Hint rom is) rcrxicolora and 
the Sphingids : — r l he horn on Hth abdominal segment: This cha¬ 
racter is as well developed in one of the North American Notodonts, 
closely allied to our A. dictaea, as in any of our British Sphingids; 
oblique stripes : these are a noticeable feature of Xotmlonta trepida. 
Shagreen or mammillary based secondary hairs occur in many butter- 
Hy lai va*, and an analogous, if not exactly similar, development is to 
be found in some Creometrid larva*. (. hara.rex jaxiux offers a good 
