Holt and Calderwood — Report on the Rarer Fishes. 407 



a condition not usually met with in percoid fishes. The air-bladder is large, 

 forming the dorsal wall of the whole of the visceral cavity. The urocyst is 

 spherical and rather thick-walled. Both specimens are immature males, the testes 

 appearing as narrow bands attached to the ventral wall of the air-bladder on either 

 side. We have not been able to detect the spleen, if that organ exists. A con- 

 siderable amount of fatty matter adheres, in the large specimens, to the 

 hsemorrhoidal vein. 



A small vein, which occupies the usual position of the lienic, appears to receive 

 part of the blood from the rete mirahile, but the greater part of the blood from that 

 structure is received by a larger and more anterior vein, which, after uniting with 

 the left spermatic, enters the Cuverian sinus. At the point of juncture of the two 

 veins last mentioned, there is in one larger specimen, an anastomosis between the two 

 venous systems, by means of a branch which leaves the portal vein just before the 

 latter splits up to enter the liver. The arterial systeai of the visceral cavity 

 presents no points of interest. 



Description of Specimens. — Scales. — The rather large, deciduous scales are 

 typically ctenoid, but very thin and semi-transparent. In a scale from the 

 neighbourhood of the lateral line about the central region of the body the 

 whole of the exposed surface is beset by fine spines, arranged in about seven 

 rows from before backwards (PI. XLii. fig. 3 a). The spines of tlie tliird and 

 sixth rows are the longest, whilst those at the extreme posterior edge are 

 the stoutest. Each spine is attached to the body of the scale for the greater 

 part of its length, having only a short free portion posteriorly resembling 

 an arrow-head in shape (fig. 3 J). The spines of the lateral line scales (fig. 3 c) 

 are few and feeble. Both examples were almost entirely denuded of scales 

 by the time they came on board,* but from those which remained, and from 

 the distinctness of the scale insertions, we believe that we have been able in fig. 3 

 to reproduce the natural condition with tolerable accuracy so far as concerns the 

 body itself. The lateral line scales were mostly in situ, but are perhaps shown in 

 the figure with too great distinctness. Whilst all the other fins were evidently 

 covered with small scales in their basal regions, the first dorsal seems to have 

 been scaleless. It is difficult to judge to what extent the head is clothed with 

 scales in the natural condition, and it is possible that the keel of the pre-operculum, 

 conspicuous in the specimens in their present state, may be really marked in life 

 by a continuous scaly integrement. 



• Lateral line. — We find fifty-three scales in our specimens. The number given 

 by Giinther in the Catalogue is forty, so that considerable variation would appear 

 to exist in this respect. 



* A result probably attributable to tbe presence of an enormous number of Spatangus Rasehii in tbe 

 net. 



