SPHINGINiE. 265 



keeping together appreciably later, but, from a Manducine standpoint, 

 they were all ancient and well-established families before Manduca was 

 heard of. Manduca is derived from some well-advanced Sphingine, and 

 is not very distantly related to some such form as Amphonyx or 

 Macrosila, the chief difference being that the plasticity in the 

 maxillary development was here very great, and permitted evolution 

 to take place in two such opposite directions as Manduca and 

 Macrosila, both, however, having a common ancestor at about the 

 developmental level of Agnus convolvuli. Manduca preserves, in 

 the pupa, a file-like rough area representing the proboscis-horn of the 

 higher Sphingids, the rest of the maxilla being, as in, them, smooth. 

 The character which may best be relied upon for the division of the 

 Sphingids (sens, restr.) from the Amorphids is the possession by 

 the Amorphids of a dorsal metathoracic suture in the pupa that 

 is functional on dehiscence. This does not occur in the Sphinges 

 (or Eumorphids). The more typical Sphinges, have, in the pupa, 

 a horn for the accommodation of the proboscis. This always 

 develops before the face-parts are pushed much upwards and back- 

 wards, so that the labrum, ceasing to be ventral, barely reaches 

 an anterior position before relief is afforded by the development of 

 the horn, and the labrum never becomes dorsal, as it does in the 

 Eumorphids, long before a horn appears. Larvally, the Amorphids 

 and the Sphingids must be associated : (i) In possessing a much 

 less primitive first stage than the Eumorphids. (2) In presenting 

 the same pattern of oblique stripes. In Eumorphids, oblique stripes 

 are not rarely developed in a reverse direction to that in which 

 they occur in Sphinges. Occasionally, however, they appear in 

 definite Sphingine manner, of which a notable example is figured 

 by Poulton (Trans. Ent. Soc. Land., 1887, pi. x., fig. 3). Our present 

 knowledge gives us apparently no ground for any theory of such 

 stripes in Eumorphids, than that they are developed quite inde- 

 pendently of any common origin with those in the Sphingine line. 

 The subdorsal line, originating at the horn of the lower forms of 

 Sphingid larvae, presents, on the 8th abdominal segment, a basis 

 for the development of an oblique stripe, that may, quite possibly, 

 therefore, arise in two different divisions of the Sphingides *. There 

 are two different groups of Sphinges, usually placed in Sphingidae, 

 whose right to be there may be seriously questioned. One of these 

 is that containing Kentrochrysalis as a central form. The larva of 

 this is said much to resemble that of Phyllosphingia dissimilis, an 

 undoubted Amorphid, and the pupa is clearly Amorphid, presenting 

 no feature in common with Hyloicus, which the imago certainly very 

 considerably resembles. Along with Kentrochrysalis would probably 

 go Sphingulus, and, perhaps, one or more species placed in Dolbina, 

 but our information about the early stages of these is too defective 

 either to support or contradict such a surmise. It may be noted 



* The Hemarid larva has rudiments of both the subdorsal and spiracular 

 series of stripes ; they are rather colour-shadings than stripes, as we use the term 

 in relation to the larvae of Amorphids and Sphingids (sens, restr.) , but requiring little 

 development to make them so; the series slope in opposite directions, the subdorsal 

 as in Sphingids and the spiracular as in the Eumorphids. In the Eumorphids the 

 stripes are only spiracular, whilst in the Amorphids and Sphingids they cover both 

 areas, but one suspects them to be of subdorsal origin (Bacot). 



