18 LÖNNBERG, FISHES FROM SPITZBERGEN AND KING CHARLES LAND. 
aberrant specimen and material kept in the Zoological State 
Museum in Stockholm. I have thus had not less than 15 
voung C. spinosus in all from Greenland and Spitzbergen 
for comparison. The result of this is that very young indi- 
viduals of this species are smooth, but already at the length 
of 14 mm., one of the specimens is beginning to develop some 
spiny tubereles on the sides. Specimens measuring more than 
20 mm. have, as a rule, a great number of spines developed 
which are arranged more or less regularly in the following 
way. One series of large spines extends from the forehead 
along the upper edge of each eye and along the curvature 
of the back where it can be traced running on the upper 
part of the caudal peduncle to the base of the caudal fin. 
Between these two, two other rows of smaller tubercles can 
be seen very plainly, beginning on the forehead and from 
there extending on each side of the base of the dorsal fins. 
The spine of each of these most median rows, which is situ- 
ated at the interspace between both dorsal fins is, at least 
in young specimens, larger and stronger than the others of 
the same row. These two pairs of spiny series correspond 
exactly with spiny rows of the same situation in Cyclopterus 
lumpus. We find in that species a series of spines from above 
the eye to the caudal fin, and another along the upper part 
of the back, best developed in the interspace between the 
dorsal fins, just as in C. spinosus, but also in many, if not 
in all specimens, conspicuous along the base of the second 
dorsal fin. In C. lumpus there are two more rows on each 
side, namely one from the opercle to the caudal fin and an- 
other from the lateral margin of the sucker to the anal fin 
and often continued by smaller spines along the base of that 
fin. To these series we find corresponding ones in OC. spino- 
sus, namely from the gill-opening a short row, which in young; 
specimens does not extend on the caudal peduncle, and an- 
other from the hind-margin of the pectoral fin to the begin- 
ning of the anal fin. But in addition to these there is in 
C. spinosus one more system of spines to which there seemsg 
at first to be no correspondence in C. lumpus, namely an 
arcuate row from the upper anterior angle of the pectoral 
fin and along the margin of this fin to the ventral row. Bu 
if this arcuate pectoral row is counted together with th 
ventral row in assuming that the ventral row in C. spinosus 
