IDIACANTHUS ANTROSTOMUS. 281 
the top of the snout, midway from the eye to the end of the snout. Four 
gills, a slit behind the fourth ; gill openings very wide, extending from the 
upper angle of the operculum down and forward to below the eye. Twelve 
short branchiostegal rays. Barbel one length of the eye from the end of 
the chin, twice as long as the head ; outer fourth of the length expanded 
into a leaf-like organ with a long point at each end, a fleshy and rather 
thick median portion and a thin transparent border. <A light organ closely 
resembling the eye in external appearance, of half the ocular length, lies 
above the middle of the upper jaw close behind a vertical from the hind 
border of the orbit ; a series of twelve similar very small organs is situated 
at the bases of the branchiostegal rays; at each side of the isthmus there 
is a row of ten of these organs; between the isthmus and the ventrals at 
each side of the median line of the belly there is a series of thirty-five and 
a little above them on each flank a parallel series; the four series continue 
backward and between ventrals and anal have twenty organs each, while 
along the side of the base of the anal there are about thirty-five more. 
Below the large organ on the cheek on the upper jaw there is a yellow 
glandular streak as long as the orbit; similar patches of glandular structures 
form three longitudinal series of blotches along the flanks, the lower of 
which le between the light organs; posteriorly along the bases of dorsal 
and anal this structure becomes more or less continuous. A prominent anal 
papilla. Vent below the twenty-seventh ray of the dorsal. 
Dorsal and anal separated from the short rays of the caudal by less than 
the orbital length; from the snout to the first ray of the dorsal nearly four 
and to the same ray of the anal more than eight lengths of the head; anal 
fin entirely within the hindmost third of the entire length; fin rays ante- 
riorly partly free, short, slender, spinous, unsegmented, but near the caudal 
becoming stouter, longer, more than twice as long as the eye, and seg- 
mented. At each side of the base of each ray the interneurals and inter- 
heemals bear a short rigid spine directed out in such a way that the two series, 
on the opposite sides of the fin, form a trough or cradle in which the rays 
lie and are protected when folded back. In the description of Jdiacanthus 
ferox it is said “Each ray starts behind a minute curved spine-like projection 
of the vertebra.” The structure thus described would interfere seriously in 
folding back the rays. Since the rays are articulated to the bones bearing 
the spines in question they would according to the statement quoted articu- 
late directly with the vertebrae. Instead of this, however, it is probable the 
