Crustacea Decapoda and Stomatopoda. 
291 
produced to a tooth at its infero-external angle. The antennal scale is nearly three 
times as long as wide and its outer margin is very slightly concave. 
The second maxillipedes are remarkable for the possession of a large protruding 
lobe, quadrate in outline, at the proximal end of the propodite. The third maxilli- 
pedes reach to the end of the antennal scale, the exopod extending beyond the end of 
the antepenultimate segment. 
In the first peraeopods (text-fig. 126) the carpus is equal in length with the palm and 
its greatest breadth is about two-thirds its extreme length ; anteriorly it is very deeply 
hollowed to receive the rounded proximal end of the chela. The second peraeopods 
(text-fig. 12c) are long and slender, reaching a little beyond the end of the scale. The 
carpus is about one and a third times the length of the chela and is between 5^ and 6 
times as long as its greatest breadth. The palm is two- thirds the length of the 
dactylus. In the third peraeopods (text-figs. i2d, e) the merus bears four spines on its 
lower margin and the carpus one near its distal end. The propodus is provided with a 
series of spinules on the same margin ; it is about 8 times as long as broad and rather 
more than 3|- times as long as the dactylus (terminal spine included). The dactylus 
bears in all 5 or 6 spines, the outermost large and strongly curved. The fifth 
peraeopods (text-figs. 12/, g) bear spines on the merus, carpus and propodus, much as 
in the case of the third pair. The propodus is from 11 to 13!^ times as long as broad and 
from 4 to 4^ times the total length of the dactylus. The latter segment bears from 29 
to 34 slender spines ; excluding these its length is a trifle more than three times its 
breadth. 
The outer uropod is provided with a series of from 18 to 21 movable spinules. 
Well-grown specimens reach a length of about 17 mm. The eggs are large 
and few in number : about 0'96 mm. by 070 mm. in longer and shorter diameter. 
Caridina serrata is allied to C. parvirostris, de Man, and C. pareparensis, de Man, but 
differs from both in the much greater proportionate length of the lateral process of the 
antennular peduncle. In addition it differs from C. parvirostris in the large size of the 
eggs and from C. pareparensis in the more deeply excavate carpus of the first pair of 
legs. InBouvier's latest scheme of classification (1913) it would come nearest to 
C. serratirostris, de Man, which it resembles in the length of the lateral process of the 
antennule. From this species, however, it differs in many respects, notably in the 
length and dentition of the rostrum and the form of the carpus in the first pair of 
legs. 
Dr. Annandale informs me that, in life, the specimens were mottled with brown- 
ish pigment and were consequently very difficult to detect on the rocks on which they 
commonly sat. They were found in pools in very small streamlets of clear water, 
devoid of weeds, on the Peak at Hong Kong, at altitudes of 12 00- 15 00 ft. The 
specimens were collected in September, three of the females being ovigerous. Two 
additional specimens from the same locality, collected by Capt. F. H.Stewart, I. M.S., 
have recently been presented to the Museum. 
Stimpson gives the habitat of his specimens as "ad insulam Hong Kong ; in 
rivulis." 
