694 Mr. R. C. L. Perkins on the 



birds as I did from fifteen to twenty years ago. Some 

 that I found commonly seem now quite extinct, and others 

 greatly reduced in number. It would be almost impossible 

 to duplicate the observations I then made. 



Nov. 10, 1911. 



What troubles me as to Hymenoptera is, that any bee 

 or wasp in life is so utterly unlike anything else, that the 

 veriest duffer of a bird can hardly mistake it for anything 

 else, and it is clear that in the islands those which remain 

 small in size with no colour of any sort (i. e. no patteru 

 and ordinary wings) are not now eaten and are fully as 

 successful as any belonging to the Colour-groups. Why 

 then on one little island (Oahu) should a lot of species 

 associate themselves in several Colour-groups for protective 

 purposes ? It would appear much more advantageous for 

 all to belong to the dominant black-coloured blue-winged 

 group on the one island, as one would say it would be 

 much easier for birds only to have to learn one colour 

 pattern than several. One tasting might do for the whole 

 lot, if they were one colour, but a number of tastings 

 might be necessary for a lot of groups ; and then I come 

 back to the old doubt, why is not the fact that all are 

 characteristically Hymenopterous (whatever be the colour) 

 sufficient in itself ? 



Nov. 15, 1911. 



If birds can select between very slight colour variations 

 so as to produce the closest mimetic resemblance, it seems 

 strange that they should not recognise a7iy Hymenopterous 

 insect as such quite apart from colour and pattern. That 

 they do recognise Hymenopterous characters other than 

 colour, seems to be proved by a mimetic Australian Man- 

 tissa. Although superficially quite unlike a Hymeno- 

 pterous insect, this Mantispa is, from its behaviour and 

 attitude, a perfect mimic — in fact the best known to me. No 

 Syrphid with all its wasp-like coloration can approach it.* 



* The mimicry of Mantispa was observed by W. M. Wheeler in 

 Nebraska (1888), G. A. K. Marshall in Natal (1896), and R. Shelford 

 in Borneo (1898-1900) and Singapore (Trans. Ent. Soc, 1902, pp. 

 536-7 ; Proc. Zool. Soc, 1902, pp. 235-7). Both Marshall and Shel- 

 ferd speak of the excellence of the mimicry on the wing. At the 

 same time Shelford's Plate (P. Z. S. 1902, XIX, tigs. 22-7), and both 

 his and Wheeler's descrii)tions show that colour may enter lai"gely 

 into the mimetic resemblance in certain species oi Mantispa, — E. B. P. 



