THE LATERAL CANAL SYSTEM. 308 
the system in Plagiostomes does not appear to have been changed to any 
extent by deep sea conditions. 
Among the bony fishes the history of the system is somewhat different. 
The arrangement of the canals differs from that on the typical Plagiostome 
to some extent, as is amply shown in the subjoined illustrations. The 
separation of the canals of the right side from those of the left is more 
general and the presence of a spiracular canal on the operculum behind the 
postorbital and usually connected with the oral is a common feature. This 
opercular canal corresponds with the spiracular (sp) of Chlamydoselachus, 
Plate LXX. In many if not most it is apparently separate from the main 
longitudinal canal at the side of the skull. Along the side of each mandible 
there is a canal which is identified with the oral (0) of the typical Galei 
(Antacea). An aural connection across the crown between the right side 
and the left may have existed in numero.s cases; if so it has to a greater 
or less degree become obsolete. A submental connection between the two 
orals, as on Chaunax, Plate LXXIII. fig. 1, appears to be exceptional. 
Accepting as typical the presence of both spiracular and postorbital, as 
seen on Lamprogrammus, Plate LXXXI. fig. 1, and on the majority of the 
figures on Plates LXXII. to LXXXIV., anomalous developments are to be 
noted on the more specialized forms, such as the changes brought about by 
a shortening of the head behind the eyes, as seen on the Scorpxnoid figured 
on Plate LXXI. fig. 1, on which the spiracular and the postorbital canals 
have fused (see Chaunax also, Pl. LXXIIL.), or on the Cottoid of the same 
Plate, fig. 2, on which the spiracular joins the orbital, forming a plan inter- 
mediate between that of the Scorpxnoid, fig. 1, and that of the Berycoid, 
fiz. 4, of the same plate. Whether the aural branches are primarily to be 
regarded as a loop, as on Lamprogrammus, Plate LXXXI. fig. 1, leaving 
the main canal opposite the parietal region and joining that canal again 
forward of the junction of the postorbital is an open question. Possibly the 
aural branches cross the head in some highly differentiated forms, like the 
Scorpenoid and the Cottoid cited above, while on others, as Lampro- 
grammus, Plate LXXXI. fig. 1, the loop represents simply another phase 
of differentiation. Owing to the amount of maceration the specimens have 
suffered it is difficult in many cases to trace the minute connecting thread 
running from disk to disk in the canals; for this reason in a number of the 
figures of specimens on which it was nearly or quite invisible no attempt 
has been made to place it; its course can safely be supplied from the posi- 
oo 
20 
