368 MR. ROBERT SIIELFORD ON 



The geographical distribution of Prosoplecta is as follows* — 

 Eight of the species are found in the Philippine Islands, one in 

 Celebes, the remaining three in Batchian and Ceram None 

 has yet been discovered in the Great Sunda Islands, so that the 

 distribution is discontinuous and serves to emphasize the view 

 that if the Philippines are To be regarded as a part of the [ndo- 

 Malayan region, their separation from adjacent land is of very 

 greal antiquity. The Philippine Islands constitute an area in 

 which insert mimicry lias attained greal perfection. That is 

 shown not only by these wonderful cockroaches mimicking 

 Coleoptera, bu1 Professor Poulton tells me thai some of the most 

 wonderful examples of mimicry amongst butterflies are known 

 from these islands only, e. g. the distasteful Danaine genus Uestia 

 is mimicked very closely by a Satyrine and an Elymniine. 

 Again, the gorgeous little Curculionidae of the genus Pachy- 

 rhynchus are mimicked by other weevils, by Longicorns, by 

 Cetoniids, and by a cricket*. A comparative study of mimetic 

 insects iii geographically adjacent but zoologically distinct areas, 

 such as Borneo, the Philippines, and Celebes, is a piece of research 

 that would surely yield some very interesting results. 



I now give a synoptical key to the genus Prosoplecta with 

 descriptions of all the species known to me. It will be observed 

 that 1 have not always Keen able to pair a species of Prosoplecta 

 with a definite Coleopterous model, hut 1 am pretty confident 



that these models will eventually he found. I have not had 



access to comprehensive collections of Philippine Coleoptera. and 

 so ii is chiefly the Philippine Prosoplectae which for tho present, 

 I am not able to match. It is unfortunate that Semper's collec- 

 tion of Philippine Coleoptera is broken up and dispersed; Stal 

 purchased bis Orthoptera for the Stockholm Museum, and here 

 it was that I found four new species of Prosoplecta, but the 

 Coleoptera captured at or about the same time and in the same 

 districts, J have not keen able to trace. 



Key to the Species of Prosoplecta. 



1. Loss convex species. Tegmina with a smooth flat- 

 teni d i nl , i -. i near the middle of the anal vein. 

 2. Pronotum piceous, margined nil round with hya- 

 line tet /'. eoccinella Sauss. 

 2'. Pronotum bright rufous 1'. bipunctata Lir. 

 1'. Very convex species. Tegmina without Battened 

 tubercles. 

 2. Ground colour of I hreous or rufous. 

 ;i. Tegmina maculate. 

 ■1. Maculae of tegmina piceous. 



5. Tegmina deeply punctate. (Pronotum with 



•1 maculae.) ..... P. trifaria Walk. 



* Semper in his ' Animal Life,' p. 390 (International Scientific Series, L890) 

 gives figures of some of these mimetic insects, amongst them one with the legend 

 " Phoraspis (grasshopper) mimics a < Joccinella." This is an error, for the Phoraspis 

 is u cockroach, apparently identical with Prosoplecta ligata Br. The species 

 described below as P. semperi is a much better mimic of the Qoccinellid fi 

 by Semper than is /'. ligata. Sempei : iced by 



Dr. Wallace in his ' I >arvi inism.' 



