292 Records of the Indian Museum. EVOL, Xai 
of the two jaws. In lineatum the upper jaw is longer, in panchax 
the two jaws are equal, in parvus the lower jaw is distinctly 
longer and the direction of the mouth-cleft in the normal unpro- 
truded condition is very similar to that in melanostigma (pl. xxv, 
figs. Trane 2). 
The upper jaw is protractile in Panchax and in the normal 
unprotruded condition a deep transverse fold is formed across the 
snout (pl. xxv, fig. 2). In Haplochilus the upper jaw is not pro- 
tractile and there is no fold. 
Teeth.—Tate Regan lays special emphasis on the difference 
iu teeth, which he believes are present in a band in Panchax, 
while they occur in a single series sometimes followed by a second 
series of minute teeth in Haplochilus. A microscopical examina- 
tion of the isolated praemaxilla and mandible of H. melanostigma 
reveals the presence of a band of teeth hardly distinguishable 
from that in species of Panchax (pl. xxv, figs. 3 and 4). 
Vomerine teeth (pl. xxv, figs. 6, 7 and 8).—These are invariably 
present in all the known Indian species of Panchax, in lineatum as a 
narrow band composed of three or four rows of teeth, in panchax in 
a single or occasionally double row; in parvus about three vestigial 
teeth are borne by the vomers at their extreme anterior end, and 
in rubrostigma and day: also they are said to occur. It is quite 
likely that all species of Panchax possess these teeth in various 
degrees of development. While they are present as a prominent 
band in lineatum, the presence of the three teeth in favvus can 
only be detected by a microscopical examination of the vomer. 
At least so far as Indian species are concerned, their presence or 
absence constitutes a safe criterion of classification. 
The pseudobranchiae are glandular, richly supplied with blood 
during life, and covered by a highly pigmented mucous membrane. 
They are present in all the three species of panchax examined and 
probably in rubrostigma and dayt, while they are absent in H. 
melanostigma. 
Gill-membranes.—These are broadly united with one another 
in H. melanostigma, while they are not united, but cleft to the 
chin in lineatum, panchax, parvus and probably in the other two 
Indian species of Panchax. 
Position of pectoral fins.—In melanostigma these are placed 
high, i.e. nearer the mid-dorsal line of the body than the mid- 
ventral, whereas in species of Panchax they are placed low 
(pl. xxv, figs. 1 and 2). 
Scales.—In meianostigma the scales possess only concentric 
(circular) striations, but in the three species examined and probably 
in all other species of Panchax radiating striae (basal radii) are pre- 
sent in addition to the concentric rings in the imbedded part of the 
scale (pl. xxv, figs. 9 and 10). 
Vertebrae.—Their number in the various species are as follows: 
in melanostigma 29, in lineatum 32, in panchax 29, in parvus 
26. ‘Tate Regan seems to have supposed that the number 29 in 
the species panchax obtains in all the species of the genus Panchax. 
