3i6 Memoirs of the Indian Museum. \\o\.. Ill, 



the more higlily specialized species have l^een able to establish themselves on all the 

 islands on one side or other of " WalUace's Line ' ' and even in a few instances {Gonatas 

 spp. and perhaps Gnuphalocnemis tridens) to establish themselves on both sides of this 

 important boundary-, it is difficult to see why more of them have not been able to 

 cross it, and why Palk Strait and Torres Strait should have formed such serious 

 obstacles to migration. 



7. APPENDIX I.— A REVISED CLASSIFICATION OF THE ACERAIINAE. 



In the key to the genera of Aceraiinae given on pp. 196-7 of the present 

 paper the genera Basilianus, Ophrygonms, and Aceraius were defined in accor- 

 dance with the views that had previously found general acceptance. Since that part 

 of the paper went to the press, however, certain new species (all of them described 

 above) have been submitted to me, which completely bridge the gaps between these 

 three genera as there defined. I have therefore been compelled to reconsider the limits 

 of the genera, and to adopt definitions more like those used in redefining the genera 

 of Gnaphalocneminae (above, pp. 199-203). Indeed, the general course of evolution 

 followed by the subfamilies Aceraiinae and Gnaphalocneminae respectively has proved 

 to be very much the same (see diagram, p. 314); and the characters by which the 

 Aceraiinae are distinguished from the Gnaphalocneminae as a whole, are probably of 

 no greater importance than those by which certain groups of the latter subfamily are 

 distinguished from one another. The majority of the genera of Gnaphalocneminae 

 were known to me only through Kuwert's descriptions until after the outlines of the 

 classification adopted above had been irrevocably fixed so far as this paper is 

 concerned. Otherwise I would have reversed the positions of the Aceraiinae 

 and Macrolininae, merging the former in the Gnaphalocneminae as the Aceraius 

 group. 



Not only do the genera Basilianus (old sense), Ophrygonius and Aceraius grade 

 oiie into another, but the genera Episphenus, Chilomazus, and Basilianus do so also; 

 for Chilomazus comptoni is even more variable than I at first supposed, and besides 

 including specimens hardly more distinctly asymmetrical than Episphenus moorei, 

 contains forms whose asymmetry is almost as great as that of the genus Basilianus, 

 which they somewhat resemble in the structure of the anterior margin of the 

 head. 



Turning now to the structure of the mandibles in these three genera, the extent to 

 which the dentition of the right mandible is reduced in any species is found to be 

 correlated with the extent of the asymmetry of the head, in all cases except that of 

 Basilianus cantori, a species in which the head is highly asymmetrical, but the 

 tnandibles scarcely more so than in . the symmetrical species Episphenus moorei. In 

 this it resembles Ophrygonius inaequalis, with which it is further connected, both 

 stucturally and zoogeographically, by certain of the new species already referred to. 

 The gaj) hitherto supposed to exist between Basilianus cantori and the genus Aceraius 

 has likewise been filled ; and the greatest gap that now remains in the Aceraiinae is 

 that between Basilianus cantori from the Himalayas and Assam, and the species from 



