( 499 ) 



alxlomen the first segment is mostly closely appressed to the thorax ; in Jfacro- 

 (/losm/m the first tergite is reduced to a very narrow strip. The fan-tail is fonud in 

 lioth sexes of a mnuber of genera, sometimes only in the c? {/'Juri/pten/x) ; it is a 

 Sesiad character indicated occasionally only by three small tnfts {Nepliele c?c?). 

 The scent-organ of the anterior coxa is sometimes strongly developed {('liromiii). 

 The mid- and liindcoxal mernm are in the greater uumlier of genera simple, or the 

 niidcoxal mernm is simply carinate or snhangnlate ; bnt tlie large sharp tootli found 

 in Si'.va and allies appears also in this tribe {Macrof/losnuni ami allies), though the 

 hindcoxal mernm is never as strongly produced as in Sesia. 



The tibiae become spinose in several genera ; the apical thorn of the foretibia 

 found in some Cephonodes {Sesiinae) and in many Spkingidae asemanophorae 

 ajipears also here, with or without an additional armature of spines {Odontosicia ; 

 J'rnserjiinus : Arctonotus ; etc.). The midtarsal comb is very often lost respect- 

 ively reduced. The spurs may be long or short, unequal or equal in length ; in the 

 latter case they are reduced. The proximal pair of the hindtibia is lost only in 

 ^firrosplhi)):(•. The pnlvillus and paronychinm are rarely lost. The mid- and hind- 

 tarsus of Macroglossum and allies is specialised in being compressed and having 

 accpiired a dense spination on the outer surface. 



The ancestral antenna was doubtless spiniform, long, and liad an elongate 

 rongh-scaled end segment. The clubbed antenna is a later ac(iuisition. Tlie short 

 end-segment found in numerous genera is a reduced long one, i.e. is a less 

 generalised character than the long segment. The length of the latter may become 

 exaggerated, as is the case in Aco.smeri/:r. The strongly clnbbed antenna has a long 

 or a short end-segment. The bristles of the end-segment are sometimes much 

 ]iroIonged, and resemble those of the Choerocainjtinae {Panacra; see also F/i Hod ila) 



The pupa with compressed tongue-case (Choerocampid tongne-case) is a 

 derivation from an anteriorly cylindrical and obtnse j)uj)a as found in Philampelicae. 

 On the other hand, the CJhoerocampid pupa of Xephclicae may become reduced, 

 assuming tlie aspect of the pupa of Sesiinne by losing the compressed projecting 

 tongue-case. Such Sesiad pupae appear frequently in genera with reduced head and 

 tongue of the imago {Deidamia, ]>ara.psa, etc.). 



The pedigree inserted below gives expression to our view of the connection 

 between the various genera. There are three main branches : one leading from 

 DeiUpliila to Darapsa ; the second from Xpphrle to Ukopalopsi/che ; the third fi-om 

 Neplidi' to De'idnmia. We thought at first that this last branch might be a 

 development from the first. But on consideration of all characters we think to be 

 correct in treating the third branch (on the left side in the ])edigree) as a derivation 

 from some such form as Nepkele instead of DeiUphila. The qualifying remark 

 ai)plied to the former pedigrees holds good also here. 



Key to the genera : 



a. Spines of first row of abdominal 



tergites not longer than broad . . . h. 

 Spines of first row of abdominal 



tergites longer than broad . . . . c. 



b. End-segment of antenna elongate, 



c? -antenna with fasciculated ciliae CL. Miicroi/lo.'f.-iiim. 

 End-segment of antenna elongate, 

 (? -antenna similar to those of ? , 

 without prolonged ciliae . . CLI. ItlMpdlopnijcht'. 



