NO. 1919. FISHES FROM JAVA— BEAN AND WEED. 589 



400 mm. The tail in the smaller of these specimens is rather longer 

 than tlescribed by Blecker/ but this character is quite variable. 



MYLIOBATIS MILVUS> MUlIer and Henle. 



One male specimen from Batavia. Length, 1,540 mm.; length 

 of disk, 820 mm.; width about 550 mm.; tail, 1,2G0 mm. 



ELOPS MACHNATA (Forskil). 



Seven specimens from Batavia, 250 to 370 mm. long, seem to rep- 

 resent this species, although they do not exactly agree with the 

 description given by Regan.^ 



We have examined 35 specimens of the genus Elops, as follows: 

 Java, 7; Ashantee,Wost Africa, 3; west coast America, 5; Australia, 

 3; east coast America, 6; Philippine Islands, 5; Hawaii, 5; Hong- 

 kong, China, 1. 



This series is entirely too limited to allow us to decide between the 

 more closely allied forms described by Regan, but the indication is 

 that our specimens will not entirely bear out his conclusions. We 

 hope to be able at some time to get a sufficient series from one locality 

 to give a clue to the amount of individual variation to expect. 



Our specimens seem to indicate separate groups as follows: (1) 

 East America, (2) West America, (3) West Africa, (4) Australia and 

 part of Philippine specimens, (5) other Pacific and Indian Ocean 

 specimens. 



We have made counts of vertebrae as follows: East coast America, 

 skeleton, 75^; Java, radiograph, 65A-; Ashantee, West Africa, radio- 

 graph, 69^; west coast America, radiograph, 79 V; Philippine Islands, 

 radiograph, 65^; Hawaii, radiograph, 68i; Hongkong, China, radio- 

 graph, 65 i. 



The species represented by our specimens from Java was described 

 b}^ Blocker imdcr the name Elops saurus.* 



MEGALOPS CYPRINOIDES (Broussonet). 



Nine specimens, 220 to 285 mm. long. Batavia. 



We have examined specimens from Australia, Samoa, and the 

 Philippine Islands. There is a possibility that the Samoan speci- 

 mens may be separable on slight grounds from the others. 



We can see no reason for separating this East Indian species gener- 

 ically from the American tarpon. Jordan and Evermann give the 

 backward insertion of the dorsal as the distinguishing character, but 

 this seems to us to be of very little value. In our specimen of M. 

 cypriiioides the dorsal is situated over the first one-third of the ven- 

 trals, wliile in atlanticus it is over the center or last onc-tliird of these 

 fins. 



1 Bleeker, Plagiostomen, p. 85, antl Day, Fish. India, p. 742. 



* Idem, p. 87. 



» Ann. Mag. Nat. Hist., ser. 8, vol. 3, 1909, pp. 37-40. 



* Atlas Ichtb., vol. 6, p. 84, pi. 2C8, fig. 3. 



