34 



MEMOIRS OF THE NATIONAL ACADEMY OF SCIENCES. 



In Citheronia, especisillj' C. regalis, we have a return to Adelocephala. with its prothoracic 

 horns and exuberant growth of spines, which are retained through all the stages of larval life. 

 The pupa, however, is more modified and diverges more widely than that of any other genus 

 from the other members of the subfamily, while the moths tend to have fairly well developed 

 maxilla', with which they lap up sweets, and sharp fore wings, anticipating the sphinges in their 

 shape. Which of these five genera gave origin to the ancestor of the Sphingidre is a question. 

 It must seem, however, as if the stem-form was an ally of Adelocephala. 



The probable course of phylogemetic development may be expressed l)y the following dia- 

 gram, which also indicates the classitication of the group: 



CrVAeronta 



Eac/es 



Syssphrnx 



rrrs 



so /a 



A defocepha7a 



IX. 



A^o/'ocfonfr</a.e(//etsroccfmpa) 



Fig. 2. — Diagram to illustrate the affinities and phylogeuy of the 

 Ceratocampinse. 



ON THE PHYLOGENY OF THE SPHINGID-ffi; THEIR DERIVATION 

 FROM THE CERATOCAMPID.<E. 



In his "Hawk moths of North America" (Bremen, Issti) the late A. R. Grote remarks on 

 the intimate relationship between the Sphingidw and Bombycidw "suggested by the American 

 group of the Ceratocampidfe," regarding the latter group as the "remains of an old type and 

 nearer to the hawk motlis than any su))family of the spinners now existing."' 



In his "Notes in ISST upon lepidopterous larva'" (Trans. Ent. Soc, London, 18S8), Professor 

 Poulton, from a study of the armature of sphingid larvse and that of A</l.ia tan, states: "We 

 have therefore an accunuiliited l)ody of facts which seem to render it certain that the Sphingidse 

 are a specialization of the group of Saturnian Bombj'ces, and that the following order represents 

 the nearest afSnity and is an approach toward the expression of genetic relationship: Sphinx, 

 Acherontia. Smerinthus, Ceratomia, Lophostethus. Aglia, Cei'atocampa (Attacus). Saturniti." He 

 adds: "The imaginal condition of the Sphingida> which come nearest to Aglia, etc., is also strongly 

 in favor of the above arrangement. They alone do not feed in the perfect state, and do not fly 

 in the cluinu'teristic manner of other hawk moths: in the strict sense of the word they are xot 

 hawk moths. Their mode of flight, and especitilly their rudimentarj' and unused mouth parts, 

 are further points of affinity to the Saturnians. 



"It therefore follows that the chief peculitii'ities of the Spliingida'. as opposed to the main 

 l)o(ly of BomVn'ces — the fact that they feed largely and are greatly speciidizcd in relation to 

 flowers — are characters which were absent from their Bombyciform ancestors, and are still absent 

 from Smerinthus, while they have been i-eacquired comparatively recently in tiie phyletic history 

 of the majority of Sphingida^. 



