1908,] Records of the Indian Museum. 271 



a little more than once and a half, and those of the 2nd pair i*8 to 

 3 times as long as the palm. The dactyli of the 3rd legs, which 

 in two of the three females appear a little shorter, but in the third 

 even a little longer than ^ of the propodite, are armed with 9 or 10 

 spines, the terminal claw included, but, according to the original 

 description, they are often armed with 11 or 12 spines. 



The specimens referred by Schenkel {I.e., p. 498) to this variety 

 seem partly to belong to another form, — probably to the var. wyckii 

 (Hicks.), — because in some of his specimens the carpus presented 

 the same shape as in the var. longirostris. 



The var. gracilipes is most closel}^ related to the var. bengalensis , 

 from which it^seems to differ by the characters of the rostrum and 

 by slightly smaller eggs. 



9. Caridina nilotica (Roux), var. minahassa, de M. 



TABI.E I. - 



(Plate XX, figs. 9, ga, gb.) 



Caridina nilotica (Roux), var. minahassa, de Man, I.e., 1902, 

 p. 895. 



Table I is taken from that in the quoted paper, the numbers 

 having been calculated. This variety is most closely related to the 

 var. longirostris , H. M. Bdw., from Oran, but differs (i) by the much 

 larger size of its eggs which are 0*55 mm. long, or once and a half 

 as long as those of the variety inhabiting the river Macta ; (2) by the 

 dactyli of the third legs being slightly shorter in proportion to the 

 length of the propodite (compare Tables B and I). The dactyli 

 of the three posterior legs have the same slender shape as those of 

 the varieties longirostris and gracilipes, but those of the third pair 

 are armed with 7 or 8 spines, the terminal claw included, whereas 

 those of the fifth pair carry 33 — 38 spinules as in the African variety. 



