HORSE-MACKERELS. 
85 
out transverse branches on each side. But in the la- 
teral line proper the transverse branches of the first 
magnitude (a pair for each plate, one upwards and 
one downwards) send out only backwards a number 
(from 5 to 7) of secondary branches", which open into 
the posterior margin of the plate. In older specimens 
of this species, in which these canals often become in- 
distinct, we may still find numerous, scattered pores, 
especially on the back and head, which are the orifices 
of the system of the lateral line. Of the other species 
of the genus, Caranx torvus has the dorsal canal simi- 
larly continued almost to the end of the second dorsal 
fin, but in Car. armatus we have not been able to trace 
it farther than the beginning of the first dorsal, and 
in some specimens of Car. Upturns it may be traced 
to the end of the second dorsal, in others only to the 
beginning of the first, as is also the case in the genus 
Selene. The variations in this respect indicate the 
transition to Temnodon and the true Mackerels, some 
of which have only a rudimentary dorsal canal, while 
others are without it altogether. 
In Caranx , as in Selene , Trachynotus and Lichia 
(I am unable to speak with certainty of the other Ca- 
rangidce), the first three interneural bones are separate 
from each other, while the others are more or less clo- 
sely united by the osseous ridges on their back and front. 
The coalescence is most complete between the 4th and 
5th. From the 4th grows out in a forward direction 
the spine mentioned in the diagnosis, and the first spi- 
nous ray in the anterior dorsal fin is supported by the 
same bone. On each of the first three interneural bones 
there is a process in a forward direction which corre- 
sponds to the spine just mentioned. These processes 
do not, however, project above the skin, though the 
tops of these three bones are sometimes visible as oblong 
knobs in the edge of the back in front of the first 
dorsal fin. 
The genus Caranx has gained some further interest 
in an anatomical respect from an observation the 
universal signification of which within the genus is, 
however, still unknown. The chief function of the air- 
bladder is, as is well known, to enable the body of the 
fish to maintain or attain the same specific gravity as 
that of the surrounding water. Thus, it is more com- 
pressed (the size of the fish smaller), but contains more 
gas (of greater density) in deep water than near the 
surface. When the fish gradually rises or sinks in the 
water or only slightly changes its position, the quantity 
of gas in the air-bladder is altered in proportion. When 
the fish is rising and it is necessary to diminish the 
quantity of gas, the Physostomi, whose air-bladder is 
furnished with a pneumatic duct, thus enjoy a great 
advantage, for those fishes which have no pneumatic 
duct, can only gradually diminish the quantity of gas 
by absorption. Hence, too, the great difference in the 
effect upon different fishes of a change in the pressure 
upon the water, e. g., when the pressure is diminished. 
Those which have no air-bladder, seem but very slightly 
affected by the change; those which have an air-bladder 
with pneumatic duct, emit gaseous bubbles through the 
mouth and gill-openings, but those which have an air- 
bladder without pneumatic duct, are drawn upwards in 
spite of their struggles to sink deeper and deeper, until, 
if they find it impossible to descend, they are quickly 
borne to the surface, where they lie swollen up and 
unable to move. In his endeavours fully to investigate 
this point Arm and Moreau'' found that Caranx trachu- 
rus , far from being borne to the surface like the Basse 
and the other Physoclysts experimented on, when the 
air above the vessel in which it was placed, was rarefied, 
seemed only slightly uneasy and emitted only small 
gaseous bubbles. On closer examination he found in 
the upper wall of the air-bladder, in a line with the 
seventh rib, an opening surrounded by a raised, val- 
vular margin and leading to a tube which follows the 
right side of the aorta in a forward direction and after 
a geniculate curve opens into the right gill -cavity. 
Another peculiarity, the universal signification of 
which is still to be decided, is the fact that the fry of 
Caranx trachurus is a parasite of the Medusae, a cir- 
cumstance which is one of the many discoveries made 
by the late Professor A. W. Malm (see below). 
The numerous species of this genus are excellent 
swimmers and thus, for the most part, have a wide 
range. They are therefore subject, as is usual under 
such circumstances, to fairly many variations of form. 
a They may be seen in McCoy’s figure {Prod. Zool. Viet., pi. XVIII), as it is reproduced in Fish. a. Fislier. N. S. Wales by 
Tenison-Woods (Sydnej T 1882) pi. XXII, fig. 1 a , showing the inside of a plate from the Australian Car. trachurus, which figure may be 
compared with our figure of the high scales in Pteri/combus brama. 
6 Ann. Sc. Nat., ser. VI, Zool., Tome IV (1876), Art. No. 8. 
