256 Cn Aquatic Carnivorous Coleoptcra or Dytiscidce. 



torms, and considerable variations in the details. Looking at these stages we find 



that as the tarsus becomes compressed laterally the two claws necessarily become 



approximated, and as the sole of the tarsus changes its direction to become a part 



of the inner face of the tarsus, so does one of the claws get drawn above the other, 



till ultimately the two claws instead of being placed side by side are situated one 



over the other, the inner claw becoming the upper one ; this inner claw which 



thus changes its position, at the same time enlarges, and becomes straighter, so 



that the two claws may be very unequal in size, it being always the upper or inner 



one that is the larger ; this condition may be seen in the Hydaticides : the lower 



claw is frequently more curved than the upper one, or in other woi'ds retains to a 



greater extent its primitive form ; and frequently there is a considerable difference 



in this respect in the sexes of one species (Ilybius, many species). The most changed 



claws are found in Hyphydrus, Laccophilus and Cybister, in all of which the 



lower claw has disappeared, and the upper one is straight (or nearly so) and 



immobile. In the genus Megadytes the two claws are present in very different 



degrees of development in the sexes of the same species ; and it has been supposed 



that the two claws are not really two but are a divided or split single claw, and 



this statement has been repeated by some of the best entomologists ; there is, 



however, no ground for the assertion, which is based on the supposition that the 



single claw of the allied genus Cybister really consists of the two claws consolidated 



into one ; a careful examination, however, of a series of species shows that this is not 



the case, but that in Cybister, where there is only one claw present, this is because the 



second claw is undeveloped or atrophied. On the single claw of Cybister there 



exists a longitudinal channel, looking something like a consolidated suture, and 



this has been cited as a proof that this single claw is a united double one ; as, 



however, where there are two claws present (in Megadytes), each of them presents 



this channel very distinctly, it is quite clear that it is no indication of the claw 



possessing it being a double one, for if so Megadytes has really four claws. 



The relative length ot the joints of the hind tarsus is rather constant in the family, 

 the rule being that the basal joint is distinctly longer than the second and that the 

 third and foui'th are each a little shorter than its predecessor, the fifth being again 

 more elongate, so as to equal or surpass in length the basal joint, but in Hyphydrus 

 the terminal joint is shorter than usual. In the Noterides, however, these 

 proportions do not hold good, and the basal joint of the tarsus is nearly or quite 

 as long as the four following joints together. 



It is worthy of remark that in the Noterides the tarsus does not, as in the other 

 Dytiscidse, develope, ■pari jxissu, with the other parts of the swimming leg, but 

 evidently is of less functional importance, so that in Hydrocanthus we find that 

 whereas the femur and tibia are more highly metamorphosed than in any other 

 Dytiscidte, the tarsus on the contrary remains small, and can be of comparatively 

 little use as a propeller. In Pelobius the legs are, as regards their structure, but 



