FISHES OP THE GULF OF MAINE 461 



rocky ground. There is no reason to suppose that the adult fish ever rise far above 

 the bottom except by accident. 



Food and habits. — Judging from the stomach contents of Scandinavian and 

 British fish (none have been examined on this side of the water) they feed 

 chiefly on shrimps, prawns, isopods, and other small crustaceans, less often on fish 

 fry. On the other hand, roclding have themselves been found in cod stomachs in 

 Massachusetts Bay, and no doubt haddock (all fish of prey, for that matter) 

 devour them. 



Rockling, like other gadoids, swim at the surface for their first few months, 

 and we have taken their pelagic fry in our tow nets at the various localities marked 

 on the accompanying chart (fig. 223) from the first week in July until October; sel- 

 dom, however, more than half a dozen in any one haul (the largest catch was 18 

 specimens). Huntsman similarly describes them as common in the center of the 

 Bay of Fundy and they have been taken in the tow at Woods Hole in April. How 

 long they remain on the surface is not known, but analogy with cod, haddock, etc., 

 suggests three months at most, and since our largest pelagic fry were 40 to 45 mm. 

 long it may be assumed they seek the bottom at a length of about 2 inches. During 

 this pelagic stage they drift with the current like other fish fry, and are at the 

 mercy of mackerel and other fish, but they are not plentiful enough in the GuK of 

 Maine to be as important an article in the diet of the mackerel as are the fry of the 

 far commoner European rocklings on the other side of the Atlantic. 



Breeding habits. — Huntsman (1922a) found the eggs of this rockling in Passama- 

 quoddy Bay in midsummer. Its breeding season probably endures from spring to 

 early autumn in the western Atlantic as it does in the eastern Atlantic,^' for Dannevig 

 (1919) records rockling eggs (probably this species) as early as the end of May near 

 Halifax, while on the other hand we have towed larvae only 5.5 mm. long as late as 

 September and October in Massachusetts Bay. Probably it spa%vns all around the 

 peripheral belt of the Gulf, with Massachusetts Bay an important nursery, judging 

 from our repeated capture of its larvae there; but our failure to find rockling larvae 

 or its pelagic fry in the central part of the Gulf, or its eggs in any of our ofl'shore 

 tows, justifies theconclusion that its breeding is limited to depths less than 75 fathoms 

 so far as the inner part of the Gulf of Maine is concerned, though it may spawn much 

 deeper on the continental slope. 



The eggs are buoyant, described (we have never seen them) as 0.66 to 0.98 mm. 

 in diameter. When newly spawned the oil is in small droplets, most of which soon 

 coalesce into one globule of 0.14 to 0.25 mm., often with one or two smaller ones 

 close to it. The danger of confusing them with squirrel-hake eggs is discussed in the 

 account of the latter species (p. 453). Newly hatched larvae are slightly more than 

 2 mm. long. The yolk is absorbed at about 3.6 mm. and the later larval stages, up 

 to about 10 mm. long, are characterized by the very large black ventral fins shown 

 in the illustrations (fig. 239c), by the presence of one post anal band of black pigment, 

 and by the short stocky body form. Young hake are more slender and have scattered 

 pigment, young cusk have two post anal bands, and all other Gulf of Maine gadoids 



3' It spawns from the end of January until August in the Baltic. 



