TJiG Bionomirs of South African Insects. 513 



All were captured at Salisbury, in Nov. 1898, with the 

 exception of those specially noted as taken between Nov. 

 1898 and Jan. 1899. 



Other species are not included in the Plate for want of 

 space. Tlius Piezia marshalli has a single median white 

 spot, of which half is on one elytron and half on the other, 

 in place of the posterior well-separated pair of spots on 

 P. selousi ; and there are other species, outlying members 

 of the group under discussion, in which the Mutilloid 

 appearance becomes less marked, while in Polyhirvia semi- 

 sitturata it fades away altogether, although this intensely 

 black beetle, with its white dorsal line anteriorly and white 

 patch posteriorly, seems to be very conspicuous. The 

 appearance of the group and the relation to the outlying 

 species suggest a strong and very complex Miillerian 

 association. The large Garahicla} of the genus Anthia 

 are either entirely black or possess a peculiar synapo- 

 sematic appearance, described on pp. 508-510, and figured 

 on PJate XVII, figs. 22 to 26. The smaller Carahidm, 

 depending upon a less development of the same defence 

 — the power of discharging a strong acid secretion — 

 have gained an appearance, due, like that common in 

 Anthia, io white markings on a black ground, but arranged 

 so as to suggest more or less strongly the likeness of a 

 Ilutilla. In a single species, as Mr. Marshall points out, 

 the resemblance is extraordinarily exact (Plate XVII, 

 fig. 11). This may be on account of habits and a mode 

 of life which render the likeness especially beneficial. 

 The smallest Carahida?,, the Atractonota and Polyhirma 

 tvnigma (Plate XVII, figs. 7-10), in shape resemble large 

 ants, and Mr. Marshall states above that the movements of 

 the former aid in producing this effect. On the other 

 hand, their white spots appear to be certainly Mutilloid or 

 perhaps rather to resemble the Mutilloid white spots of 

 the other unant-like CarahidcV. Nothing is more char- 

 acteristic of a Miillerian (synaposematic) group than the 

 complexity of likeness which is thus revealed, and yet in 

 the light of the great hypothesis which we owe to Fritz 

 Mllller it is not difficult to understand the general principles 

 which account for its existence. 



The GarabidX' are a powerful, specially defended group, 

 and it is of advantage to be recognized as belonging to 

 the group, even though it is no doubt of still greater 

 advantage to be mistaken, as may happen at a distance, 



TRANS. ENT. SOC. LOND. 1902. — PART III. (NOV.) 34 



