568 Notes in 1887 upon lepidopterous larve, ce. 
The points (1), (2), (6), and (7), are evidently adapta- 
tions to the peculiar conditions of the larva, and cannot 
be considered to prove any great divergence in affinity, 
any more than the various adaptations which form 
such sharp characteristics within the group of Sphingide 
themselves (such as the terrifying marks of Cherocampa, 
&c.). In the present state of our knowledge, we cannot 
venture upon an opinion as to the meaning of points 
(5) and (9). So small a number of larval ontogenies 
have been carefully worked through, that it is impos- 
sible to decide whether such characters are adaptive. 
The subdorsal line (4) may be represented in part, as I 
have already suggested. Points (8) and (8) in reality 
indicate affinity quite as much as divergence. Thus the 
caudal horn degenerates in size and shape in the later 
stages of many Sphinx larve. It nearly disappears in 
the last stage of Cherocampa porcellus; it is absent 
from all except the earliest stages (if indeed it is present 
in these latter) of Deilephila vespertilio and Pterogon 
cenothere (Weismann, I. c., pp. 209 and 259). The 
ontogeny of Agha is more exaggerated than that of 
any Sphinx larva. It commences with amore specialised 
caudal horn than that which any Sphinx possesses, and 
associated with a specialised remnant of the spinous 
covering of allied Bombyx larve ; and suddenly at the 
last ecdysis, all these prominent features are lost, as 
completely as the horn is lost in the later stages of 
certain Sphinx larvee. 
The character indicated in (8) is of extreme interest, — 
not, indeed, as showing affinity with the Sphingide,— 
but with allied Bombyx larve, with which the Sphingide 
are thus brought into association through Aglia. This rela- 
tionship will be discussed in the next paragraph. 
3. THE NATURAL POSITION OF THE SPHINGIDA.—The most 
interesting question raised by the ontogeny of Aglia tau 
is that of the position of the Sphingide. I have enume- 
rated a large number of important characters, in which 
this larva is related to the Sphingide and especially to 
the genus Smerinthus. These characters are so numerous, 
that it is, | think, impossible to explain them as due to 
recent convergence caused by adaptive changes. The 
pronounced condition of some of the most striking 
