262 MEMOIRS NATIONAL ACADEMY OF SCIENCES. vol. xii, 



crameri is evidently the most extreme and recent form of Attacus, and an islandic offshoot of 



the genus. 



COSCINOCERA HERCULES (Miskin). 



Plate LXXXIV, fig. 1; LXXXV; LXXXVI. 



Attacus hercules Miskin [Proc. Ent. Soc. Loud., 1875, p. XXVI]. — Oberthur, Etudes d'entomologie, liv. XIX, p. 34, 

 PI. 1, c? , fig. 1, 1894. 



[Watson, Wild Silk Moths of the World (Manchester, 1912), PI. I, gives good colored figures 

 of both sexes.] 



Geographical distribution. — [Australia] ; Ansus, Jobi Island, northern New Guinea (Doherty, 



Oberthur). 



ATTACUS Linne. 

 [Attacus Linne, Syst. Nat., I (1767), p. 809.] 



[Mr. Watson contributes the following notes on Attacus: 



The group of genera which at one time comprised the genus Attacus have lately been 

 rightly separated in the different zoographical regions of the globe into various genera, and no 

 doubt even more will be done in this way in the near future with other large genera as life 

 histories have been studied and worked out. As at present constituted the genus Attacus is 

 composed of two small groups of genera forming two sections; these are Attacus proper, of which 

 the type is atlas (L.), with its many geographical races or subspecies, and of which I now de- 

 scribe a new subspecies from the Andaman Islands, and the small group which I call Archeo- 

 attacus, the type of which is A. edwardsi (Wht.) . If we compare the representatives of the differ- 

 ent Asian genera together in both sexes, such as A. atlas (L.), A. lorquini (Feld.), A. edwardsi 

 (Wht.), Coscinocera Tiercules (Misk.), and Drepanoptera vacuna (West), it will be seen there are 

 two distinct types here represented. It is likely that Attacus is a derivate from the Actios 

 group of genera through Argema, Coscinocera, and A. lorquini, the female of which with its long 

 drawn out hind wing is seen to approach Coscinocera. It will be noticed the type of coloring 

 and general appearance of A. atlas, lorquini, dohertyi (Roths.), and Coscinocera hercules are all 

 [of] one character and separated easily from Drepanoptera vacuna, Archeoattacus edwardsi, and 

 Philosamia walkeri (Feld.), which last few genera form a little group with general resemblances 

 peculiar to themselves. 



It is not beyond impossibility that the simdarity of the Attacus group, including Coscino- 

 cera, to the edwardsi group is a case of Miillerian adaptation; the strong pungent scent given off 

 by A. edwardsi and from which I can tell fully 24 hours before a moth emerges, a scent which 

 is not given off to my knowledge by any of the many subspecies of atlas which I have hatched, 

 but given off by Philosamia, may have something to do with it. 



Rothschild's plate of A. staudingeri in Novitates Zoologicae, 1895, Plate 10, figure 2, shows 

 the forked costa, the less highly pectinated antennae, the postdiscal white fascia curving out- 

 wards towards the apex, and the distinctive variations between the width of the subcostal 

 veins. I have not seen A. staudingeri, but the illustration on Plate 10 is evidently perfectly 

 trustworthy, the characters there found agreeing accurately with edwardsi. The plate shows 

 the two forms of antennae side by side; it figures A. dohertyi and Archeoattacus staudingeri. — 

 J. H. Watson.] 



ATTACUS ATLAS Linng. 



Plate XXVI, fig. 1; XL VI, fig. 3; LXXXVIII, fig. t; LXXXIX; XC; XCI, fig. a. 



[Bombyx atlas Linne, Syst. Nat. (1758), I, p. 495.] 



In its venation and in the antennae and other characters, imaginal and larval, the Asiatic 

 species of Attacus differ markedly from those of the new world, or Neogaea, and should be 

 separated generically. If we are in doubt as to the generic value of the adult features, the 

 larvae are certainly very different, and can not be included in the same category with the Neogaeic 

 caterpillars. It was not until I had decided that the Asiatic forms were generically different 

 from the American species, that I met with the suggestion of Sonthonnax. [This was written 

 before Grote proposed the name Rothschildia for the American forms.] 



