406 Annals Entomological Society of America [Vol. XI, 



(Fig. 8), or Darapsa the maxillae are too long to occupy the 

 space between the labrum and the caudal margin of the wings 

 if they extended straight caudad, so the proximal part of the 

 maxillas arches away from the body and forms a prominent 

 curve near the cephalic end. In other genera as Pholus, (Fig. 7), 

 Celerio, Xylophanes, etc., the head has been tipped back so 

 the labrum appears on the dorsal surface. This gives some 

 additional length, and, as the wings extend about three-fifths 

 the length of the body, and the segments are lengthened ven- 

 trally to accommodate these, it usually is sufficient for the 

 length of the maxilla. In cases where it is not, the proximal 

 part arches out from the surface of the body as previously 

 described. This is true of Xylophanes tersa. There are still 

 some pupae, however, with considerably longer maxillae, and the 

 method of taking care of this extra length seems to me to have 

 developed from those previously described. It would be rather 

 difficult for the maxillae to be arched very far from the surface, 

 so the extra length is taken up in a loop at the proximal end. 

 This loop is, in most instances, closely appressed to the ventral 

 surface of the body. In others, as in the common tomato worm 

 pupa (Fig. 3), it is strongly arched, while in Herse cingulata the 

 loop is equal in length to the remainder and is recurved (Fig. 2). 

 When one considers that the maxilla is double in the maxillary 

 loop (Fig. 1) it will be seen that in such genera as Herse, the 

 maxillae must be at least one and one-half times the length of 

 the body. The maxillary loop has been referred to by some 

 authors as the raised tongue case, or the jug-handle tongue case, 

 as contrasted to the ordinary type of maxilla (Fig. 5) which 

 was said to be "sunken." Other authors referred to it simply 

 as the tongue case, ignoring the remainder of the maxillse. In 

 all measurements involving the maxillae, or the length of wings 

 when comparison is made, the length is always measured from 

 the suture separating the labrum and maxillae. 



CLASSIFICATION. 



Strange as it may seem, the spiracular furrows and the 

 maxillae have proven to be the best characters for separating 

 the pupae into the various divisions of Rothschild and Jordan. 

 In their "Revision of the Lepidopterous Family Sphingidae" in 

 1903 they divide the family into two groups, the AsemanaphorcB 



