190 Memoirs of the Indian Museum. [Voi,. IV, 



very variable in form. The lateral portions are grooved longitudinally throughout the 

 length of the somite. On the sixth somite there are four large tubercles placed in a 

 transverse series. In de Man's type specimen those of the submedian pair bear each 

 two stout spinules, but ten were observed by Fukuda in a larger example from the 

 Japanese coast. The lateral tubercles, which may also bear a varying number of 

 spinules, are sometimes broken up into a number of smaller eminences. Numerous 

 stout spinules also occur between these and the lateral edges, and in the Japanese ex- 

 ample several small spinules are found on the anterior margin. 



The telson is about as long as wide and the deep median excavation extends to al- 

 most half its entire length. It bears three median bosses : the median being triangular 

 and pointed behind, the lateral pyriform, posteriorly rounded and not reaching the distal 

 margin. The surface of the telson, like that of the sixth somite, is beset with numerous 

 spinules, but their precise number seems subject to much variation. The median boss 

 bears nine in the type specimen, while on those of the lateral pair there are ten. In 

 Fukuda' s specimen there are on the contrary about thirty on each of the bosses. The 

 lateral margin is armed with six to fourteen spinules, and there may be as many as 

 twenty others between this edge and the lateral dorsal bosses. On the margins of the 

 large distal excavation there are six to twelve pairs of spinules. 



There is a stout spinule at the proximal end of the basal segment of the uropods and 

 the inner spine of the broad ventral bifurcate process is about half the length of the outer. 

 The inner segment of the exopod bears ten movable spines on its external margin. 



The variation in the number of spinules in this species is very remarkable. It 

 may be shown eventually that Fukuda' s example deserves varietal recognition ; but, 

 inasmuch as only three specimens are yet known, it is impossible at present to reach 

 any definite conclusion. 



De Man notes the presence of bright green colouration on the tubercles of the last 

 two segments, but Fukuda did not find this in the specimen which he records, and ex- 

 amined when living. 



There are no examples of Gonodactylus brooksi in the Indian Museum. The two 

 type specimens, the largest 28 mm. in length, were found at Pulo Edam, Java (de Man), 

 while Fukuda's specimen, 39-5 mm., was found in Tateyama Bay, Awa, Japan. 



23, Gonodactylus hystrix (Nobili). 



1899. Protosqmlla hystrix, Nobili, Ann. Mus. Civ. Genova (2), XX, p. 276. 



This species, which has not been figured, is evidently very closely allied to G. 

 brooksi. According to the published account it differs from the latter in having fifteen 

 carinae on the fifth abdominal somite (including the two marginal), the seven median 

 being limited to its posterior half. The sixth somite is covered with small spinules 

 with recurved points, directed backwards, and arranged in three regular series. The 

 telson is beset with spinules of a similar character, and the three dorsal bosses are of 

 a circular shape. 



Nobili records a single specimen, 15 mm. in length, from Beagle Bay, British New 

 Guinea. 



