168 



INTRODUCTION. 



very large ; tibial spurs large ; abdomen with the apex uncovered, 

 in the first group produced into a strong style, visible segments Jive or 

 six, abdominal surface convex or subcarinate. 



We have regarded the BhipiphoridtE as distinct, although they 

 might be with reason included under this family. About 400 

 Mordellidje are known ; they are variable, but easily recog- 

 nizable and fall into two subfamilies as follows : — 



I. Apex of abdomen produced into a strong style .... Mordellinje. 

 II. Apex of abdomen not produced into a style Anaspin^s. 



The perfect insects are found for the most part on flowers 

 herbage, shrubs, and low branches of trees, but some occur in 

 decaying trunks. The larvae are in some cases found in rotten 

 wood, while in others they live and feed in the stems of plants ; 

 sometimes they are found in the old burrows of wood-boring 

 insects. The larvae of Mordellistena are elongate and curved, and 

 are furnished (at all events in some species) with protuberances 

 on the first five or six abdominal segments ; these are also present 

 in the pupae ; the anal segment is variable, and in Anaspis is cleft 

 into two processes. A few are believed to be carnivorous, but 

 this does not appear as yet to have been decisively proved. 



The Anaspiis.e have been regarded as peculiar to northern 

 temperate climates, while the Mordellinje have been recorded 

 chiefly from Europe and North America ; but Mr. Guy Marshall 

 informs me that he has eight species of the former and sixteen of 

 the latter from South Africa. Mordella and Mordellistena are 

 represented in Ceylon, if the records of Walker and Motschulsky 

 are correct. 



Family 84. RHIPIPHOBIME. 



Head vertical, resting against the anterior coxce ; antennce eleven- 

 or ten-jointed, variable and varying in the sexes ; prothorax as broad 

 at base as elytra, with the sides not forming a sharp edge; anterior 

 coxce large, conical, contiguous, cavities widely open behind and con- 

 fluent ; posterior coxce transverse, more or less contiguous ; elytra 

 sometimes covering the abdomen, often dehiscent, and rarely wanting 

 (in the larviform female of Rhipidius); metasternum large : abdo- 

 men ivith free segments, varying in number ; claws variable, pectinate, 

 toothed, or (rarely) simple. 



There are no clear characters at present defined for the sepa- 

 ration of this family from the MoRDELLiDyE, and, as Evaniocera 

 and its allies form a transitional group between the two, it might 

 perhaps be better to follow Dr. Sharp in regarding them as 

 merely subfamilies of one family ; at the same time, as he himself 

 hints, it is possible that a study of the head may cause the sepa- 

 ration of the group into several families, so that it can hardly be 



