304 Records of the Indian Museum. |Vol. XIII, 



ig to 32 teeth' (usually 22 to 31) of which i or 2, rarely 3, are 

 placed on the carapace behind the orbital notch. On the lower 

 border there are from i to 14 teeth (usually 2 to 9) ; the distal 

 third of the lower margin is in most cases unar med. 



The lateral process of the antennular peduncle sometimes 

 reaches only to the end of the basal segment, in other cases to 

 about one-third the length of the second segment. 



The carpus of the first peraeopods is comparatively slender, 

 from 2"o (Lake Torrens) to 2*9 times as long as broad and is some- 

 times, as shown in text- fig, 5^, much less deeply excavate than in 

 other species. The carpus of the second pair (text-fig. 56) is from 

 5*8 to 7"5 times as long as broad. The propodus of the third pair 

 (text-fig. 5c) is from 3*5 to 4*0^ times as long as the dactylus (ter- 

 minal spine included). The dactylus (text-fig. 5^) bears from 9 to 

 13 spines, usually 9 to ir; excluding these its length is from 3*0 

 to 3*6 times its breadth. In the fifth peraeopods (text-figs. $e, f) 

 the propodus is from 30 to (rarely) 3-8 times the length of the 

 dactylus. The latter segment, spinules excluded, is ver^^ variable 

 in form, from 3*3 to nearly 5 ^ times as long as wide. The spinules 

 vary in number from 28 to 82.* 



The third and fourth legs of the male show no signs of sexual 

 modification. 



No ovigerous females are present in the material examined. 

 The largest of the vSydney specimens is 27 mm. in length ; an in- 

 dividual from "S. Australian waters" is rather larger, about 31 

 mm. 



The specimens examined are from Clj^'de, near Sj^dney, from 

 Lake Torrens in S. Australia and from " S. Australian waters." 

 The first of these samples includes the type specimens ^ which bear 

 the number 7590-2/10 in the Zoological Survey register. The 

 specimens recorded by Ortmann from Burnett in Queensland, by 

 Bouvier from Melbourne and by Thomson from Victoria and New 

 South Wales are presumably to be referred to this species. The 

 identit}^ of von Martens' examples from Adenare near Flores is 

 quite uncertain. 



The material I have examined shows an unusually great range 

 of variation and it is possible, as noted above, that more than one 

 definable race of the species exists in Australia ; the specimens in 

 my hands are, however, not sufficiently numerous to afford evi- 

 dence that this is really the case. 



^ The rostral formulae in the three samples are as follows : — In 12 specs, from 

 Sydney "^.1! ■ in 6 specs, from I ,ake Torrens ?^^^ : in 5 specs, from " S. Austra- 

 ian waters " ' j"^- . 



2 In a female from Lake Torrens. 



3 3-3 to 4'o in most cases. The specimen with a proportion of nearly- 5 is 

 perhaps an abnormality. 



* From 28 to 65 in the Sydney specimens. 



6 Owing to a very unfortunate accident the types have baen destroyed since 

 the description was drawn up. The only portions of them that remain are cer- 

 tain appendages mounted on slides for microscopic examination. 



