66 
In Ladmeides, the large size of the supra-spiracular warts on the 
thoracic segments of many species is probably also a sign of the dual 
character of iii in the stirps. 
With ('itheronia, probably, and certainly with A gl ia tan, the 
character and meaning of the thoracic horns or processes, is, I take it, 
entirely different from those of the Saturniids proper. With A. tau, 
so far as my observation goes, the anterior and posterior dorsal 
thoracic tubercles do not combine to join the horns, those on the 
dorsum of the meta-thorax being developments of i alone, while the 
foremost pair of horns is placed on the pro-thorax, and not the meso- 
thorax, as with Satumia ( H . io). I believe that this difference also 
applies to Citheronia, but have not yet had larvae for examination. 
With Dimurpha versicolor a, the meso- and meta-thoracic tubercles 
are in a single row, and consist of warts bearing several hairs, probably 
i and ii dorsal, iii is a wart of about equal size, and the wart below the 
spiracle is probably iv and v. 
I regret to say that my notes on Bombgx mori were made long 
before 1 realised the value of the difference between thoracic and 
abdominal tubercles, and the hairs on the only preserved larvte (in 1st 
instar), that I possess, are too confused to enable one to judge of their 
position. 
I trust that these few notes on the position of the dorsal meso- and 
meta-thoracic tubercles will enable my readers to form some idea of 
the probable value of this character as a means of separating the 
Sphingids into two main divisions: (a) in which the larvar : have the 
dorsal tubercles on these segments set as anterior and posterior, i and 
ii being on separate sub-segments; and (If) in which they are on the 
same sub-segment, i and ii having their basis conjoined, or the two 
setie being on a single plate. 
Classification on this character, it will be seen, throws the 
Eumorphid (Choerocampa ) and Phri/xid (Deilephila) groups into A, 
while Sji/un.r, Ainurphids, Hemaridx, Sesiids (Macrogloxxidx) fall into 
group J> — in contrast to the more usual grouping in which the 
Awor/diids are treated as one group, and the swift-flying and feeding 
species as another. 
There is, however, a feature peculiar to Amorphid larvie which 
distinguishes them from larvae of any of the other groups known to 
me. I refer to the coat of secondary hairsl which is present in the 
1st stadium of Amorphid larvae, but which does not appear until 
after the first moult in larvae of the other groups. Although this 
feature is a far more striking one than that upon which 1 have based 
my divisions A and 11, it is to my thinking of much less importance, 
consisting at best of the acquirement of a secondary specialisation at 
an earlier stage, whereas the position of the dorsal tubercles on 
meso- and meta-thoracic, although apparent only in the 1st 
stadium, is in no way altered, so far as 1 have been able to observe, 
by subsequent moults, but ceases to be noticeable, owing to the 
reduction in size of the primary setie and the consequent difficulty of 
* They must be examined in 1st stadium, as after the 1st moult the primary 
hairs degenerate and become obscured by secondary characters. 
t lieferred to as Shagreen hairs by Prof. E. B. Poulton, and Mammillary hairs 
by Dr. T. A. Chapman, on account of their raised bases, which form so characteristic 
a feature of some larvee, especially those of Smerinthus. 
