Crustacea Decapoda and Stomatopoda. 279 
only in two races of the species,' viz. the typical form, which is found in Egypt, 
and the subspecies- paucipara from Natal. From both these forms the subspecies 
macrophora is distinguished by the greater proportionate length of the dactylus of the 
third legs, while from paucipara it also differs in the smaller number of spinules on 
the dactylus of the last leg. 
C. n. macrophora may also be distinguished from all the other known races by 
the reduced number of teeth on the rostrum, a feature which is especially marked in 
the case of those on the lower border. 
I have Httle doubt that the two mutilated specimens recorded by I^anchester ^ 
from the River Petwi, Tale Sap, as Caridina wycki are to be referred to this sub- 
species. 
The specimens were all obtained in January, 1916, at the northern end of the 
Tale Sap in and near the mouth of the Patalung River. The water where they were 
found was quite fresh, though subject to slight alterations of level according to the 
state of the tide, and probably remains fresh throughout the year. The types bear 
the number 9664/10 in the register of the Zoological Survey of India. 
Caridina brachydactyla, de Man. 
1892. Caridina wyckii, de Man {nec Hicksori), in Weber's Zool. Ergebn. Reise Nied. Ost-Ind., 
II, p. 386, pi. xxiv, figs. 29 /, g, i, a, k, cc, dd. 
1908. Caridina nilotica var. brachydactyla, de Man, Rec. Ind. Mus., II, p. 269. 
1913. Caridina brachydactyla, Bouvier, Trans. Linn. Soc, Zool. (2), XV, pp. 463, 466. 
.subsp. peninsularis, nov. 
A number of specimens collected by Dr. Annandale near Patani, in the Siamese 
Malay States and on Penang I. appear to represent a local race of deMan's C. nilotica var. 
brachydactyla. This form, hitherto known only from Celebes, Flores and Saleyer, dif- 
fers notably from all other varieties of C. nilotica in the very short dactyli of the last 
three pairs of legs and Bouvier, whom I follow, has recently given it full specific rank. 
Minor points of distinction are to be found between individuals from Patani and 
those from Penang, while the specimens from both these localities in my opinion differ 
sufficiently from those described by de Man to justify their separation as a distinct 
subspecies. 
The rostrum (text-fig. ma) always exceeds the length of the antennular peduncle 
and in some cases extends a trifle beyond the end of the antennal scale. It is a little 
upturned distally, more rarely straight, and is armed above with a series of 21 to 37 
(usually 25 to 32) teeth of which 3 or 4 (usually 3) are situated on the carapace behind 
the orbital notch. In most of the specimens examined by de Man a considerable length 
of the rostrum towards its distal end is unarmed, except for the presence of from i to 3 
subterminal teeth, in this respect resembling C. nilotica. In the specimens before me 
the condition is quite different. The teeth, in the great majority of cases, stretch un- 
1 The size of the eggs in C, mHoHcci subsp. wycki, Hickson, a race found in L/ake Tondano in Celebes, is at present 
unknown. No ovigerous females occur among cotypes of the subspecies preserved in the Indian Museum, 
2 I/anchester, Proc. Zool. Soc. London, 1901, p. 560. 
