RELATIONSHIP OF E. VERSICOLOR TO THE SPHINGIDES. 247 



E. versicolor and C. dpcnnr, though not perhaps in itself sufficient to 

 establish the relationship, seems to greatly strengthen the case when 

 taken in conjunction with the similarities I have already mentioned 

 in the larval state. It would be interesting to know if similar habits 

 or pupal spines occur in other species of the Sphin<jidaf. 



Even in the imagines, I fancy I can detect a certain amount of 

 resemblance. In shape, the wings of E. versicolor dilier chiefly from 

 those of Smerint Jills in having a more regular outline and greater 

 breadth, but when Ave consider the width of the wings in S. populi, 

 compared with those of S. tiliae, and the regularity of outline in those 

 of their near ally, S. lifjustri, we see that these difi'erences are not of 

 great importance. Again, the scheme of coloration is not unlike, from 

 a general point of view, though I must admit that there is considerable 

 diflerence in detail. There is also the discoidal lunule on the fore- 

 wings, which we find white in S. oceUatus and S. pnjndi, and black in 

 S. lijiustri. A careful examination of the ocellated spots on the fore- 

 wings of S. carpini also shows this lunule faintly visible, and it can be 

 traced in L. quercifolia ; but as we find a white spot in just about the same 

 position in Bomhyx qncrcus and Odonestis potatoria, and a well-marked 

 dark lunule in many species of Liparis, not to mention its occurrence in 

 Notodonta and some Noctuids, it is probably too general a character to 

 be of much value in tracing a connection between E. versicolor and the 

 Sphingides. On the whole, I do not trust much to these rather far- 

 fetched resemblances in the imago, and certainly shall not qixarrel 

 with anyone who difi'ers from me on these points. 



To sum up : After my examination of the larvae last May, I had 

 pretty well decided in my own mind that a relationship existed 

 between E. versicolor and Smerinthus, and an examination of the 

 pupte, coupled with Mr. Buckler's remarks on its habits and their 

 similarity to those I had observed in C. elpenor, strengthened the 

 impression. Any doubts I had left were swept away by a perusal 

 of Mr. Poulton's paper " On the Ontogeny of Acilia tau " (Trans. Ent. 

 Soc. Land., 18S8). This species, unlike E. versicolor, has important 

 characters in common with Snierinthus in its first stage, and after the 

 first moult Mr. Poulton remarks that the young larva has a growth 

 of shagreen dots exactly like those of SnierintJms and Sphinx, in 

 which genera these structures also became prominent in the second 

 stage ; and he also remarks, " That a further resemblance is shown in 

 the fact that each shagreen tubercle is terminated by a short bristle, with 

 a clavate end." Speaking of the oblique stripes which in A. tan, as in 

 E. versicolor, slope in the opposite direction to those of Swerinthus, he 

 says, " Too much importance must not be attached to the fact that the 

 stripes slope in the reverse direction to that commonly occurring in the 

 SphiiKjidae ,iov nndouhted Sphin.rlAvwee [e.fi.,thegeni\s8esia(Macroiil()ssa)] 

 possess similarly reversed stripes." Mr. Poulton considers the natural 

 position of the Spliiniiidac and allied groups to be as follows : Sphinx, 

 Acherontio, Snierinthus, Ceratonia, Lophostethus, Af/lia, Attacus, 

 Saturnia ; and remarks that Endromis will be found to occupy a 

 position in this list. My idea is that E. versicolor would come some- 

 where between Snierinthus and Saturnia, but having no acquaintance 

 with the genera Mr. Poulton places betAveen them, I cannot venture 

 on anything definite as to its position in the list. It seems also to be 

 related to ('hoerocampa (elpenor), but in what degree I am not pre- 



