— 45 — .1(11 IS. SCHMIDT 



only souii iiiiilo a low nut very well preserved specimens of the very earliest developmental 

 stages (ca. 4 — 5 mm.) and I iind it necessary, now that I have myself taken and studied 

 well-preserved specimens, to explain that the schematic figure mentioned does not give 

 a correct picture of the normal appearance. I give here on p. 46 a new schematic 

 ligure based on some quite young specimens (S'/a— 4 mm.) from the Skager Kak^. It will 

 be seen from this that the postanal pigment consists of a continuous dorsal streak and 

 a similarly continuous ventral streak. 



These streaks must be considered as produced by the uniting of the two anterior pig- 

 ment bars (III and II), which occur in the other Gadus species of the 1st group (provided 

 with a mediolateral streak, see Gadus monograph, I, p. 8 and 28), G. saïda, callarias and 

 virens, and it thus appears that in Gadus pollachius, even in the very earliest stages, we 

 do not find the postanal pigment divided into tranverse bars as is typical for the species 

 of this group. Whether this may be the case in the embryonic stages or the very earliest 

 larval stages, I do not know, but 1 am not inclined to believe so from the appearance of 

 postlarval developmental stages only S'/'z— 3-^/i mm. long. Gadus pollachius must therefore 

 be regarded as being with respect to the pigmentation in the earliest stages the least pri- 

 mitive of the species of the 1st group. The two species, G. saïda and callarias, are the 

 most primitive, all the bars I, II, III being present. Gadus virens has only the bars II 

 and 111, but they remain separated for a long time, and in Gadus pollachms, where like- 

 wise only II and III are present, the postanal pigmentation in the earliest stages may be 

 said to consist of II -\- III (see fig. 3). 



Of the pigment bar No. I, present near the end of the tail in Gadus callarias and 

 saïda, there is not the slightest trace in G. pollachius and the two postanal streaks end 

 abruptly posteriorly as in G. virens. 



From this fact that the young of the pollack never has separate pigment bars, they 

 are more easily distinguished from those of G. virens (and callarias) in the very earliest 

 stages than 1 believed when I wrote my Gadus monograph. 



In addition to what has been said in the monograph, p. 17 — 21, where for the rest 

 everything is correct regarding the appearance of the fry with exception of the account of 

 the above mentioned youngest specimen 3^/* mm. long, I wish to mention here that the 

 mediolateral streak is lacking in the very youngest stages 3^2 — 4 mm. long. This only 

 begins to appear posteriorly at a length of scarcely 5 mm., but already at quite 6 mm. 

 it may reach forwards nearly or quite to the anus. 



The ventral preanal pigment is already well-developed in the earliest stages (in con- 

 trast to G. virens) and the same applies to the postanal pigment. Both the dorsal and 

 ventral streaks of the latter consist of a double row of well-marked stellate chromatophores. 

 Posteriorly these streaks extend to an equal distance or the ventral is a trifle longer than 

 the dorsal. They reach somewhat further back than in Gadus virens, but end abruptly as 

 in that species, so that the colourless end of the tail stands out in sharp contrast to 

 the rest. 



The ventral streak reaches to the anus anteriorly and the dorsal streak is continued 

 almost without break into the occipital pigment (in contrast to Gadus virens). 



' I may remark that all transitions are found in the same sample, between the earliest postlarval 

 stages only 3^l2 mm. long and the older specimens already provided with a mediolateral line. 



