230 Records of the Indian Museum. [Vol. XIII, 



the females are small, about o'56 mm. by 0*43 mm. in longer and 

 shorter diameter. 



Living specimens are transparent, the dark gastric and hepatic 

 masses being as a rule clearly visible through the carapace. In 

 large individuals reddish flecks and suffusions are sometimes found 

 on the sides of the abdomen and the postero-dorsal margin of each 

 somite is rather deeply tinged with the same colour. The eggs 

 are pale greenish yellow. Small specimens are quite colourless. 



This remarkable species shows in a new and very striking man- 

 ner the close relation that exists between the genera Leander and 

 Palaemon, and once again awakens doubts as to whether our classi- 

 fication is correct. 



The only essential difference between the two genera rests in 

 the hepatic spine, which is present in Palaemon and absent in Lean- 

 der , and, in comparing normal forms of the latter genus with spe- 

 cies of Palaemon in which the chelipedes of the male have not 

 assumed a peculiar development, it is frequently by this point alone 

 that the two genera can be distinguished. The value of the charac- 

 ter has recently been much discounted by Caiman's discovery that 

 it is not, as was previously thought, absolutely constant.' In Palae- 

 mon hildebrandti, a form which is restricted to Madagascar, the 

 hepatic tooth ma}' be either present or absent. In all other res- 

 pects this species is a typical Palaemon; it shows no affinity with 

 Leander and cannot be regarded as establishing a link between 

 the two genera. It indicates none the less that the hepatic tooth 

 may occasionally prove an unreliable factor. 



The existence of such a form as Palaemon mirabilis is both 

 unexpected and perplexing, for, except for the presence of the 

 hepatic tooth, its affinities seem to be unmistakably with Lean- 

 der stylijerus and its allies, a group of species which form an out- 

 standing and apparently highly specialized section of the genus. 

 Were it not for the tooth in question P. mirabilis would undoubt- 

 edly be given a place in this section of Leander, diff'erring from L. 

 stylijerus merely in the abrupt curtailment of the rostrum and in 

 the proportionate length of the various segments of the legs. 

 Moreover, so far as I am aware, the species bears no resemblance 

 to any Palaemon hitherto described. 



We see, therefore, that if the character of the hepatic tooth 

 be upheld as a generic determinant, a double relationship can be 

 traced between the two genera: firstly, through the unspecialized 

 forms of each and secondly, — if my interpretation of the facts be 

 correct — between Palaemon mirabilis and the specialized Leander 

 of the stylijerus-gxon^. If these relationships are accepted as in- 

 dications of the course which evolution has taken, as I think they 

 must be, we are forced to admit the existence of a double line of 

 descent — which is manifestly impossible in a rational scheme of 

 classification. It should be noted that the styliferiis-gxonp does 

 not appear to be a disconnected entity, such as might have evolved 



^ Caiman, Proc, Zool. Soc. Loudon, 1913, p. 928. 



