AY, On Aquatic Carnivorous Coleoptera or Dytiscide. 
ES 
different condition to what we find existing in the predaceous Coleoptera, yet 
there is reason to think that the coxee of these beetles are now smaller than they 
were formerly, and that the condition of these parts in the insects we now call 
Carabidee and Dytiscide may in far distant periods have been more similar to 
what still exists in Blatta than is at present the case, and this will afford a clue to 
the modes of growth that have preceded the present structural conditions : and 
ultimately enable us to determine which of these are higher, which lower. 
The Haliplidee were formerly included among the Dytiscide, but their separation 
was suggested by Leconte, and has been efiected by C. J. Thomson. Schaum in- 
cluded them again in the Dytiscidee, remarking however (Insect. Deutsch. I, pt. 2, 
pp. 9 and 10), “that their legs are not swimming legs in shape, but only by 
ciliation, and that they differ so much from the true Dytiscidee in the insertion of 
the antennze and the number of their joints that they could be erected into a 
peculiar family, were it not tor the fact that Pelobius existed to unite the two 
groups together.” ‘This latter remark of Schaum's I fail to understand: There 
is no sense in which Pelobius can be correctly said to unite (or be a connecting link 
between) the Haliplide and Dytiscide. It is true that it is doubtful whether 
Pelobius should be included in the Dytiseidee at all, but it does not approximate 
in the slightest degree to any of the special peculiarities of the Haliplide ; and it 
does not follow that because neither of them is Dytiscidee therefore they are allied. 
To include the Haliplide in one family with the Dytiscidze, while these latter are 
kept separate from the Carabidee is certainly an erroneous course, for the Haliplidee 
not only fail to possess the peculiar coxa of Dytiscidze, but have in fact that part 
modified from the Carabideous type or structure in a totally different direction 
from what the Dytiscide have, and cannot therefore be classed with these. It is 
true they exhibit the important peculiarity of glabrous antenne, and that they 
share this in common with the Dytiscide, but this, although it may be a sufficient 
reason for separating them from the Carabide, is not of itself enough to warrant 
their union with the Dytiscide. While should the glabrous antenne be considered 
insufficient (when this part has been sufficiently studied in the Carabide), to 
warrant their isolation, then they must be classed as a group of Carabidee,* but not 
with the Dytiscidee. 
The Dytiscidee show but little approximation to any other beetles, besides those 
already alluded to. The other two families of water beetles, Gyrinide and 
Hydrophilide, are so distinct that no one in later times thinks of classifying 
them together. 
* I may here notice, though foreign to my immediate subject, that the Haliplide differ from the 
Carabide, in the structure of the front of the head, and in the insertion of the antenns, and approach in 
these respects to the Cicindelide. 
