Sasaki: On Pikk i-rom Sakiiai.in and Liidiis nu/unU (Dvuowski) 53 



this a'si tallies with Dybowski's statement. 



The teeth arc slender, sharply pointed, usually more or less curved crcs- 

 centwise, the larger ones are of almost canine appearance. There are about 

 50 teeth on each of the dcntariesa, of which those on the anterior one-third 

 of the bone arc equally small and arrange themselves in a row, while on the 

 remaining part tlicy are very unequal in size, and form tvio or three irregular 

 rows. Of the latter teeth several in the outside arc pnjniincnt and become 

 regularly smaller as they receed backwards, these teeth arc \isible before dis- 

 section. On each intcnnaxillary the teeth number about 15, forming a regular 

 row, and are equally minute. Each [lalatine bone has roughly 170 teeth arrang- 

 ed in several rows, which become longer anteriorly and iatcrually ; of these ten 

 or more near the antero-intcrnal margin of the bone arc especially large. The 

 vomerine teeth are about 130 in number; they become longer anteriorly, and 

 about ten on the anterior end of the bone are very long. Th.e glcssoliyal 

 has many teeth arranged in several rows on the posterior half. 



The dorsal and anal fins are opposite each other in front of the caudal 

 peduncle. The dorsal fin is a little longer tlian the anal, and originates a 

 little more anteriorly. The rays of the dorsal fin, inclusive of the rudimentary 

 ones in the anterior which are thix-e or four in number, number 19 or 20, 

 while the anal fin rays, also including the rudimentary ones, vary from 16 to 

 iS. As regards these numbei^s again there can be found an agreement with 

 the description by Dybowski. though they are both of greater number than 

 given by Pallas, who mentions the rays to number 14 in each of the dorsal 

 and anal. A .'■imilar dificren.ce from Pallas' description e.\ists also in the 

 number of the pectoral fin rays, which are 14 or iS in the specimens before 

 mc as well as in the case of Dybowski and not 1 2 as Pallas gives. These 

 differences arc the only discernible dissimmilarities from Pallas' description ; 

 but this, perhaps, is due to the difference in counting and discounting of the 

 rudimentary rays. 



The scales of a longitudinal row along the mid-lateral line of the body 

 range from 137 to 149 in number, counted from tlie scale nc.\t to the gill 

 opening to that opposite to the last vertebra. Dybow.ski describes them to be 



