Packard.] 
104 
[May 7, 
there are in the same stages two dorsal spines on the eighth ab- 
dominal segment, like those on segments 1-7. 
In Splungicampa bicolor , whose larva is more closely allied to 
Eacles and Cithelonia than to Anisota and Dryocampa, 4 here is a 
large bright red, spiny horn on the eighth abdominal segment in all 
the stages from the time of hatching. In stage i, there is a median 
double piliferous wart on the ninth segment, which in stage n be- 
comes a very short double spine. In stage i the eighth uromeral 
spine, as we may call it, is forked at the tip, bearing two bristles, 
but in the succeeding stages the tip is single, and this suggests that 
while the eighth uromeral spine may be two-bristled or even 
slightly forked at the end, the spine itself is single, and not de- 
rived from a coalesced pair of spines ; and this may have been the 
case with the ancestors of the Sphingidae. Of course it may be 
possible that in such a case as Sphingicampa, and in the Sphingi- 
dae, the spine may have in embryonic life been double, and that the 
coalescence occurred before birth. 
In Eacles imperialis of the first stage, both my specimens and 
Mr. Bridgham’s excellent drawings show that the six great dorsal 
thoracic spines are deeply forked, while there is a similar but 
slightly smaller median single forked spine on the eighth abdomi- 
nal segment, each fork bearing a seta. There is also a similar single 
dorsal ninth uromeral spine, but nearly one-third smaller. In the 
fully grown larva the eighth uromeral spine is single and ends in a 
single stout spine. 
In Citheronia regalis, the arrangement of spines is the same as in 
Eacles, only the shape of the spines differing, the proportion being 
about the same. The arrangement of the spines of the end of the body 
are also the same both in C. regalis and C. sepulcralis, judging from 
alcoholic specimens. In the mature larva the “caudal spine” is 
large and single, and behind it on the ninth segment is a small very 
short one, like the three others on each side of it on the same seg- 
ment. It will be seen from the facts we have recorded that in some 
cases W. Muller’s view as to the double origin seems to be sup- 
ported, while there are other facts opposed to the view. There is, 
however, no doubt but that the caudal horn in the Attacidae and 
Ceratocampidae is an individualized spiniferous tubercle, which has 
been eliminated and specialized from those on the same and other 
abdominal segments, and that this process may have occurred in 
any group of Bombyces, in which it is present in the mature larva. 
