The fishes of the Danmark Expedition. 647 



became completely frozen in the formaline in the course of the 

 winter, so that their form, fin-rays etc. were spoilt in part. From 

 the living specimens, however, I made the following observations 

 (larvae of 1907; PI. XLVI, fig. 11). 



The anus lies somewhat in front of the middle, the distance 

 from anus to end of the tail being 12 mm and from anus to snout 

 8 mm. The pectorals are very large, reaching to the anus, and have 

 fully developed rays; the first dorsal is distinct from the second, 

 which is the longest and highest; the rounded caudal is connected 

 by a membrane with the second dorsal and anal fins; the latter is 

 just as broad as D~ and reaches forward right to the anus; posteriorly 

 it ends a little before D^. All these fins have rays (which could 

 not be counted with certainty however), whereas the ventral fins 

 are only present as rudiments. There are 2 distinct spines in the 

 occipital region (the hindmost pair in the adult), but neither nasal 

 nor frontal spines are to be seen; on the other hand the curved 

 row of rudimentary spines are distinct on the preoperculum. The 

 root of the tail is distinctly bent upwards, and the lower, plate-like 

 hypural elements are well-developed. 



The larvae from 1906 differed from the foregoing in that 

 they did not have the larval form, and the rudiments of the spines, 

 pigmentation, fins etc. were more developed. The specimen of 19 mm 

 (PI. XLVI, fig. 12) had a characteristic form, which I may call cod-like, 

 to mark the stage between the larval form and the adult. Its pectorals 

 are still larger than in the 3 foregoing and the ventrals are more 

 distinct and the unpaired fins more sharply marked off from one 

 another (though the connecting membrane is still present). The 

 spiny equipment of the head is almost as well-developed as in the 

 specimen of 20 mm. The hypural bones by their growth have made 

 the base of the tail more square-shaped posteriorly. — The specimen of 

 20 mm (PI. XLVI, fig. 13) greatly resembles that figured by Ad. S.Jensen. 

 The head is flattened above and large in proportion to the trunk; 

 the whole form of the fish is now distinctly "scorpion-like". The 

 two occipital spines are now large and distinct and the 2 frontal 

 and the 2 nasal spines are indicated; the preopercular row is also 

 distinct'. The rays in the unpaired fins are now very distinct, but 

 the fin-membrane between A, С and D- is still present. The number 

 of rays in the 2 larvae are as follows: 



D^ i)2 A P 



6(7) 11(12) 14 16 19 mm. 



7 13 14 16 20 — 



^ Comparing the 5 larvae it is seen that the spines of the liead are developed in the 

 following order: occipital, then preopercular and lastly frontal and nasal spines. 

 XLV. 45 



