Ld 
On Aquatic Carnivorous Coleoptera or Dytiscide. 971 
the Dytisci fragmentati have but little connection with the Dytisci complicati (only 
the variable swimming leg and bare antennze), that the distinctions between the 
Carabide and Dytiscidse are not very important, and that on the other hand 
the Dytisci complicati are distinguished by a character absolutely peculiar to them 
among the whole enormous mass of Coleoptera, the question arises would it 
not be a more natural classification to follow the structure of the coxal cavity, 
treating it as of superior importance to the hind coxa and swimming leg ; and thus 
massing together the Dytiscidee and Carabidze into one whole comprising three series 
determined by the number of pieces entering into the structure of the coxal cavities. 
I think it is probable that this would be an improvement, and content myself with 
pointing it out to naturalists who may be inclined to discuss the very difficult 
question of the classification of the families of Coleoptera. There is no harm in 
mentioning that in the Mammalia the water-frequenting members are not classed 
together because of the similarity in their forms and in their apparatus for locomotion, 
but are distributed in different families on account of other peculiarities. 
If this view were adopted, and the modifications in the structure of the middle 
coxal cavities were considered as of predominant importance over the structure of 
the hind coxe, we would, before long, have to discuss as to which of the three 
forms of structure of the middle coxal cavities is the higher and which the lower. 
The facts as generalized are very peculiar; thus we tind that the greater part of 
the Carabidee have the coxal cavities formed by only two pieces ; that the smaller 
first or fragmentary series of the Carabidee, together with the first or smaller and 
fragmentary series of the Dytiscidee, and the whole (?) of the Cicindelidze have 
these cavities formed by three pieces; while the greater part of the mass of the 
Dytiscidee, including the higher forms of the family, have four pieces entering into 
the formation of these parts. Now, as the Cicindelidx are, on the whole, the 
highest of the families of predaceous Coleoptera, as they have fewer pieces to form 
the cavities than have the Dytisci complicati, it would appear that cavities formed 
by a small number of pieces were characteristic of a higher grade of organization 
than those with more pieces in their composition ; but against this must be set the 
tact, that the higher Dytiscidee have more pieces in these cavities than have the 
lower Dytiscide. It would thus appear that however important may be the 
structure of these cavities as a means of classification, it is not at present in 
relation with general perfection of the organization. And the explanation of the 
anomaly may probably be found in the fact that the mode of formation of these 
parts was determined at a very early period of the development, and when the 
present gradations as to higher and lower did not exist. That the primitive con- 
dition of these parts of the thorax and legs may possibly have been extremely 
different to what we find it at present, may be inferred by an examination of 
Biatta in the Orthoptera, where we find the coxz to be enormous in size, while 
the thoracic pieces are correspondingly reduced ; now, although this is a very 
TRANS. ROY, DUB. SOC., N.S. VOL. IL. 61 
