520 PROCEEDINGS OF THE NATIONAL MUSEUM. vol. xxxi. 



between base of ventral and vent. In CoUlcJithy.s fragilis there are 

 17 or 18 in this region. 



Color in spirits, pinkish and yellowish, slightly dusky above, iins 

 without color. On ventral surface there are rows of creamy white 

 spots, each one indicating a scale. A median row of 1 forward from 

 between ventrals to isthmus; a row of 3 each side of this from each 

 ventral forward; 10 or 11 from each ventral l)ack to vent; 21 from 

 front of anal along base of anal and lower part of caudal peduncle to 

 base of lower caudal rays. The rows extending upward on side of 

 belly to a level with lower pectoral rays. In CoJl/cJithys fragilis 

 there are 8 spots forward from between ventrals in a median line; 6 or 

 7 on each side of this; 17 or 18 from each ventral to side of vent; 27 or 

 28 from front of ventral to base of caudal. The vertical distance 

 between the rows is decreased in a corresponding degree. 



The above description is from the tN'pe, 110 mm. in length. Other 

 specimens vary as follows: Depth 3 to 3^^ in length. Eye li to If in 

 head. Dorsal, VIII, 23 to IX, 25; anal, II, 11 or II, 12. 



Besides the differences noted above, this species has a larger e3'e 

 (from 5 to 6 in head in CoJ I ! cht Ji ys frag ill x) and a shorter soft dorsal 

 (26 or 27 rays in CoU/chthys fragilis). The number of spots on the 

 ventral surface is the most conspicuous character for separating these 

 two species. Collichfhys lucidus has more dorsal rays than either of 

 these. 



. About thirty specimens were collected, the largest 150 mm. in 

 length. 



Type.^Cat. No. 55632, U.S.N.M.; cotypes No. 9884, Stanford 

 University. 



15. PSEUDOSCI^NA MITSUKURII (Jordan and Snyder). 



A specimen 31 cm. in length agrees ver}" well with the description 

 of the type and with a specimen of about equal size from Matsushima, 

 Japan. 



For the present we place this species in Pseadoscipena rather than in 

 PseudotoUthus. The relations of both genera to allied forms needs 

 further study. 



Family TRIGLID^. 



16. LEPIDOTRIGLA MICROPTERA Giinther. 



Five specimens taken at Port Arthur. They differ from specimens 

 from Japan in having a wider, flatter interorbital space; the rostral 

 processes more projecting, the distance across them greater, and armed 

 with longer, stronger spines. The Port Arthur specimens, tliougb 

 differing considerably in the extreme from the Japanese specimens, 

 grade into the latter in all of these characters, as is shown in the 

 following table of measurements: 



