1869.] MR. ST. GEORGE MIVART ON ANUROUS BATRACHIANS. 283 
web is.sometimes so variable in one and the same genus, that, I snb- 
mit, family distinctness can hardly be determined by it. I therefore 
propose to treat the presence or absence of disks as a character of 
minor importance, and to place in one family all forms not otherwise 
distinguishable than by the presence or absence of a web. 
The two remarkable and aberrant genera Pipa and Dactylethra 
(which agree in having the eustachian tubes so united as to have but 
one pharyngeal orifice, and in being destitute of a tongue) I am not 
disposed to regard as low forms. The absence of the tongue is a 
-defect ; but then the highest of all reptiles (the Crocodilia) differ from 
the inferior forms of their class by a similar one. The single eusta- 
chian opening is certainly rather an evidence of higher development. 
It is with considerable doubt, and mainly out of deference to the 
opinion of others, that | make Pipa and Dactylethra types of one 
and the same primary subordinal division, It may be that Pipa is 
to the edentulous Anura what Dactylethra is to the toothed forms, 
and that they are, as Mr. Cope has suggested*, extremes of two 
different series. Nevertheless the agreement between these genera 
is very remarkable, not only as to the absence of a tongue and the 
structure of the tympanum, but also as regards the expanded sacral 
vertebra, the absence of parotoids and of teeth on the palate, the 
hidden tympanum, and the broadly webbed toes. 
In searching for an important character by which to divide pri- 
marily the rest of the Anura, I have to choose between that afforded 
by the expansion or non-expansion of the transverse process of the 
sacral vertebra, and the presence or absence of maxillary teeth, as 
(considering the conditions presented by the genus Callula) I de- 
cline altogether to adopt as an important distinction the expansion 
of the ends of the digits into disks. 
The sacral dilatation is certainly a very remarkable structure, and 
one probably rather essential than adaptive+. Nevertheless, were 
-it to be selected as the primary character, it would lead to the asso- 
ciation of Engystoma and Bufo with Hyla and Discoglossus, and the 
radical separation of the last-named genus from Rana, and of Poly- 
pedates and Hylodes from Hyla. 1 therefore conclude to adopt that 
character which Dr. Giinther considers the most important after the 
* Natural History Review, 1865, vol. v. p. 98. 
+ Dr. Ginther says, ‘‘ What influence the dilated or more cylindrical form of 
the’ process of sacral vertebra has upon the mode of life is difficult to explain, as 
it does not absolutely correspond with other physiological or anatomical charac- 
ters. The pelvis obtains by the dilatation of these processes much more firm- 
ness ; and a lateral motion is more or less entirely impeded. We should there- 
fore expect to meet with this character in those Batrachians which are provided 
with the longest and most: powerful hind legs, according to the physical problem 
that the longest lever requires the strongest centre. But, on the contrary, the 
Batrachians with the shortest legs exhibit a much-dilated sacral vertebra, whilst 
on the other hand the long-legged Hylide show the same peculiarity. In fact, 
this osteological form seems to be connected with none of the modifications of loco- 
motion; for we find among these Batrachians good swimmers as well as bad, 
tree- as well as earth-frogs, those which always hop as well as those which often 
crawl ; finally we are unable to refer to it a peculiarity of any part of the propa- 
gation.” (P. Z. 8. 1858, p. 340.) 
