MANDUCIDI. 393 



Tribe : Manducidi. 



It has been generally acknowledged among lepidopterists that 

 the Manducids were closely allied to the Sphingids (sens, strict.), but 

 this has not prevented (see anted, iii., pp. 355-357) various authors 

 from placing them in a separate subfamily, and such, at first, was our 

 intention (loc. cit., p. 367), but further study (anted,, iv., pp. 264-5) nas 

 led us to a different conclusion, and, in spite of the great modification 

 observed in the pupa and imago, we are now inclined to consider 

 the group only of tribal value in the subfamily Sphingiuae (anted, 

 p. 263). We have already stated (loc. cit., pp. 264, 273) that the 

 Agriids (Agrius convolvuli) and the Sphingids (Sphinx ligustri) were 

 not, after all, so very closely allied, and we have placed (pp. 273, 

 296, 329) their British representatives in two different tribes — Agriidi 

 and Sphingidi — in spite of the fact that Meyrick and many other 

 authors place such widely different species in the same genus, Sphinx, 

 but we were not altogether prepared for Rothschild and Jordan's 

 marvellous innovation (Revision of the Sphingidae, pp. 4 et seq.), by 

 means of which they unite our Agriidi (under the name of Herse), 

 Manducidi (under the name of Acheroniia), Megacorma [pbliqua) and 

 Coelonia (fulvinotata) under the name Acherontiicae, and separate 

 them from the rest of our tribes in Sphinginae (anted, p. 273), 

 which they retain under the name Sphingicae, so that, by this 

 arrangement, the Cocytiids and Phlegethontiids (with the exception 

 of Coelonia) are brought into close association with the Sphingidi and 

 Hyloicidi, and are widely separated from the Agriids and Manducids. 

 With this separation, up to a point, we agree, although we are 

 not sufficiently informed to support Rothschild and Jordan's sweeping 

 statement {Revision, p. 6) that the species of Agrius (Herse) are " all 

 closely allied with one another in structure and pattern. They show 

 close affinities with Coelonia and Acherontia, and have nothing to do 

 with Protoparce (rustica) nor with Phlegethontius ( Cocytius)." Whilst 

 recognising the necessity for a wider separation of the Sphingids, 

 Agriids, Phlegethontiids, Cocytiids, &c. (anted, p. 273), than 

 has previously been allowed, we still think the Manducids should 

 be dissociated tribally from the Agriids, the Manducids being 

 specialised ovally and larvally in the direction of the Sphingidi rather 

 than in that of the Agriidi, and in the pupa and imago are 

 quite separate in their specialisations, the structure of the 

 antennae, maxillae, legs and abdomen, being quite unique, whilst the 

 wing-shape and scaling are supposed by some authors to be 

 specialised rather in the direction of the Amorphids. It may be 

 well here to note Rothschild and Jordan's argument for uniting the 

 Agriids and Manducids. They write (Revision of the Sphingidae, p. 5J: 

 " This small tribe is a derivation from the following one (Sphingicae), 

 with which it is closely connected through Coelonia (fulvinotata) and 

 Xanthopan (morgani). The relationship between Acherontia and the 

 genera with which it is here united in one group has never been 

 noted before. Acherontia occupies, in the classification of all authors, 

 quite an isolated position. Herse (convolvuli) and Coelonia (fulvinotata) 

 have been considered generically identical either with Protoparce or 

 with Hyloicus, and the Sphingid, here called Megacorma (obliqua), as 

 a near ally of the common oriental menephron, standing in Kirby's 

 Catalogue under Meganoton. When studying the structure of the species 



