On Aquatic Carnivorous Coleoptera or DytiscldcB. i)51 



(Dytiscus dorsalis, Celina, Sternopriscus, Hydroporus ublitus, e.(j.), ox may be very 

 small, as in Dytiscus ovatus, (Hyphydrus, No. 347), and numerous Hydropori. The 

 reduction of the pieces of the prosternura seems to attain its maximum in Hyphydrus, 

 where the anterior and inter-coxal bands form a mere slender frame for the setting 

 of the coxa?, which are longer than is usual in the other Hydroporides. The pro- 

 sternal process is not only variable in size but is exti'emely so in form ; its junction 

 with the inter-coxal band is frequently very conspicuous and it may project at the 

 point of junction as a kind of angular or tubercle-like prominence ; this is especially 

 the case when the prosternal process is thick and powerful as in Pachydrus, where 

 the prominence alluded to is clearly seen to be the result of a thickening and growth 

 of the prosternal process, while no corresponding development of the inter-coxal 

 band has taken place ; hence at the point between the front coxa3 where occurs the 

 junction of the thickened prosternal process with the feeble inter-coxal band, there 

 is an angular prominence. It is worthy of remark that the New World seems to 

 have been more favourable to the production of this thickening of the prosternal 

 process than the Old World has been ; not only does it reach its maximum in the 

 New World Pachydrus, but it will be found to occur with great frequency in North 

 A.merican Hydropori, and indeed in certain cases {e.g., Nos. 492 and 493, Deronectes 

 prosternalis, and Dytiscus griseostriatus) is almost the only character by which a 

 New World species is distinguished from an Old World one. 



In short though the members of the Irlydroporides agree among themselves in 

 showing a feebleness of the anterior parts of the prosternum, yet that feebleness is 

 accompanied by great variety of structure, and I do not see the least reason for 

 considering it as evidence of genetic connection among the insects possessing it. 

 Its functional value will probably be expressed in terms defining the nature of the 

 movements of the head and front legs performed by the insects. 



The front and middle tarsi of the Hydroporides show great variety, although 

 agreeing in the generalized and guarded definition given above. Moreover they 

 show much difference in the sexes of the same species, and this renders their use 

 as a means of definition somewhat more difficult. Thus though the thi-ee basal 

 joints agree in being more or less fiat, and in presenting beneath a sole clothed 

 with pubescence, yet as will be seen by examining the sexes of Dytiscus ovatus 

 (No. 347), this structure is much more conspicuous in the males than in the 

 females, for in the former the three joints are comparatively broad and flat, and 

 present a well marked sole beneath, while in the female they are compressed in such 

 a way as to be extremely convex transversely on their upper face and to present 

 but a small surface or sole beneath. It is the rule throughout the aggregtite that 

 the males have the front and middle tarsi broader, and so displaying the character- 

 istic structure more markedly than the females. Although the male tarsi may be said 

 to have the basal joints dilated, they never have any tendency to assume the beautiful 

 patellate form occurring in the corresponding sex of the Hydaticides, Dytiscini and 



