Batrachia. j-^^ 
specimens from Cochin and northern Tenasserim that I believe to be identical with 
Ferguson's larva, thougli the}' do not altogether agree with his description and 
figures. An outline of the mouth-disc of a specimen from Cochin is given here and a 
detailed drawing of the same individual is reproduced on pi. vi. 
The following specimen is preserved in the collection of the Indian Museum 
15747 • • ■ • Travaiicorc . . P.ritish IMuscum (Kx.). 
THE MUTUAL RELATIONSHIP BETWEEN THE FROGS OF 
THE RANA TIGRINA AND R. LIMNOCHARIS GROUPS. 
At first sight, if only the adult frogs were examined, it might seem that these 
two groups were quite distinct, if closely allied, R. wasl being in a way a connectant 
form but actually allied to R. limnocharis rather than to R. tigrina. The larvae how- 
ever, so far as they are known, apparently fall not into two, but into three different 
groups and the lines of cleavage do not coincide with those that separate the 
adults. 
Unfortunately the larva of R. wasl is unknown, but those of R. tigrina and R. 
nigulosa , on the one hand, differ in important characters from those of R. limnocharis 
and R. cancrivora on the other. The identity of the tadpole of R. hrevipalmata is 
still doubtful, but, if the provisional identification here suggested should prove cor- 
rect, it differs considerably from those of all the other species. The larva of R. limno- 
charis nilagirica agrees in essential features with that of the typical form of the species, 
but differs in minor characters. 
The tadpole of R. limnocharis is by no means highly specialized. Its mouth-parts 
are of normal form and, in particular, bear a number of rows of teeth that seem 
to be common in the genus, viz. five. That of R. cancrivora , if we accept van Kam- 
pen's identification, only differs from that of R. limnocharis in colouration and pro- 
portions, features which are both variable. The tadpole of R. tigrina , on the other 
hand, has mouth-parts of a rather highly specialized kind, and differs, so far as I 
know, from all other tadpoles as yet described except that of R. rugnlosa in the horny 
armature of the interior of the mouth; it has also an unusually powerful beak and 
numerous rows of teeth on the mouth-disc. All of these are probably adaptations 
for an active predaceous life. The supposed tadpole of R. hrevipalmata , on the other 
hand, has fewer rows of teeth than is perhaps normal in the genus, viz. three; its beak 
is rather feeble and its mouth unarmed internally. Unfortunately we know nothing 
of its habits, but it probably feeds on minute organisms or decayed matter. 
A*, hrevipalmata must be dismissed from further consideration in this connection 
owing to lack of precise information, but there are two interesting points in the adult 
frogs of R. tigrina and R. cancrivora which seem to bear out the view that they are 
convergent forms, less closely connected with one another by descent than R. tigrina 
and R. rugulosa. On the feet of R. cancrivora there is, as has already been pointed 
out, no internal metatarsal tubercle and the internal metatarsal fold is perhaps a 
little less well-developed than is normal in R. tigrina. At the proximal end of this 
