968 On Aquatic Carnivorous Coleoptera or Dytiscide. 
between the two internal lamine of the hind cox ; 8, the ciliation of the posterior 
tibiz and tarsi; and 9, the change of form in the hind legs by which they are 
adapted for swimming. Of these characters, No. 1 is quite as marked in many 
Pseudomorphini of the Carabidee and in Cyclosomus as it is in many Dytiscidze 
and more so than ina few Dytiscide. No. 2, the structure of the Dytiscid head 
is greatly approximated by the Pseudomorphini. 38, the antennz in several 
members of the Pseudomorphini, Scaritini, and Trachypachini are more or less 
completely deprived of sensitive pubescence; in other cases (as in Anthia) the 
exserted sete are but little developed. 4, a prosternal process always exists in 
the Dytiscide, but is also found in some Carabide, especially Cyclosomus, 
Trachypachini and Omophron. 5, the prosternal process in the Dytiscidze does 
not always reach the metasternum, it fails to do so in the Vatellini, in Tyndalhy- 
drus, and in Andex ; while on the other hand in Cyclosomus of the Carabidew, the 
prosternal process reaches the metasternum. 6, the external lamina of the hind 
cox is always much larger in the Dytiscidze than it is in the Carabide, it varies, 
however, extremely in size in the former family, so that from this character alone it 
would not be easy to say where the line of demarcation between the two families 
should be drawn. 7, this, like the preceding character, is constant in the Dytiscidee, 
but the hind coxze are likewise co-adapted in some Carabidze (Trachypachini and 
other members of the first series of Carabide), so that here again the Dytiscid 
character is only an exaggeration of what it is in some Carabide. 8, the hind 
tibize and tarsi are in some Dytiscide but feebly cillated, and in some members of 
the Scaritini group of Carabidze the hind tibize are strongly ciliate externally. 9, 
in some of the Dytiscide there is so little difference from the Carabidz in the 
form of the swimming leg, that the character cannot be relied on as diagnostic of 
the family. 
We see then that the relations between the two families are very intimate, so 
intimate that it may be, I believe, stated with correctness, that the Dytiscide are 
modified Carabide. By which I mean to say that there is reason to believe that 
the remote ancestors of the Dytiscidz, had before they inhabited the waters 
acquired to a considerable extent an organization similar in many respects to that 
which some Carabide still possess. In the genus Pelobius we have clearly an 
altered Carabid; and the frequent persistence in other Dytiscide of traces of 
a Carabid-like structure of the hind tarsus, and other similar facts makes one 
believe that what is clearly true of Pelobius, is probably true of other Dytiscidee. 
Herr Kolbe in the Deutsche Entomologische Zeitschrift, 1880, pp. 258, et seq., 
has published a new, and from many points, very interesting classification of the 
Carabide, Dytiscidee, Cicindelide, Haliplhdz and Gyrinide, based on the theory 
that the land beetles are descendants of the water beetles. This is, however, 
most certainly an erroneous conclusion ; it is, as 1 have said above, possible to 
believe that the Dytiscide are modified Carabidee, but it is quite incredible to me 
