248 BULLETIN OF THE BUREAU OF FISHERIES 



The eggs are buoyant, transparent, spherical, and 0.7 to 0.8 mm. in diameter. 

 There is usually a single oil globule of about 0.17 to 0.2 mm. In newly spawned 

 eggs, however, there may be two globules, which coalesce as development advances. 13 

 At a temperature of 65° — the summer state of the surface of Massachusetts Bay — 

 incubation occupies less than 48 hours. It is probable that development can only 

 proceed in comparatively warm water, though the lower temperature limit to 

 successful reproduction is not known. The larvsB are about 2 mm. long at hatching, 

 comparatively stout, with the vent situated far forward on one side and considerably 

 above the margin of the finfold, and with large black chromatophores scattered 

 over head and trunk. By the third day after hatching, when the larva is about 

 2.3 mm. long, the yolk is absorbed and the pigment has gathered in four charac- 

 teristic patches — one on the nape, one in the dorsal region of the abdominal cavity, 

 one on the dorsal side and one on the ventral side of the trunk behind the vent. 

 Dorsal, anal, and caudal fin rays are visible in larvae of 6 mm., when the body has 

 already begun to assume the deep compressed form so characteristic of the adult 

 butterfish. At a length of 15 mm. the caudal fin is deeply forked, the unpaired 

 fins are formed, and the little fish resembles the adult sufficiently for ready 

 identification." 



During their first summer young butterfish, like young haddock, often live in 

 the shelter of the larger jellyfishes, and Goode (1888, p. 222) graphically described 

 the fry of 2 to 2 J^ inches as swimming among the tentacles of Cyanea (10 or 15 

 little fish under one Medusa), where they find protection from larger fish but to 

 which they sometimes fall prey. This association, however, is not essential to 

 their welfare, for fry are often seen living independently at the surface, particularly 

 in sheltered bays west and south of Cape Cod, and we have found no young butterfish 

 with the many Cyanea that we have captured in the Gulf of Maine. 



It seems that the fry that are hatched earliest in the season grow to a length 

 of 3 to 4 inches by autumn, great numbers of that size having been taken in Rhode 

 Island waters in October, but late-hatched fish are probably not more than 2 to 3 

 inches long at the beginning of winter, and they can grow little during the cold 

 season, for little fish of 3 to 5 inches are seen again in the spring. A series of 

 measurements made by Welsh at Atlantic City, N. J., in August, 1921, throws 

 some light on the subsequent rate of growth. The fish fell into two groups — one 

 ranging from 4 to 5}4 (and averaging about 4%) inches and the other from 7}4 

 to 10 J^ inches. Probably those of the first group (which were much the more 

 numerous) were in their second summer and those of the second size group in 

 their third or perhaps fourth summer. These measurements suggest, furthermore, 

 that some may mature when 1 year old and that all do so when 2 years old. 



Butterfish are caught in pounds, traps, weirs, a few in gill nets, seines, and 

 otter trawls. We have never heard of one biting a hook. 



13 A large series of butterfish eggs artificially fertilized at Gloucester hatchery have been available for comparison with the 

 pelagic eggs taken in the tow nets. 



" Kuntz and Eadclirle (1918, pp. 112-116, figs. 58-68) give a full account of the embryology and larval development of the 

 butterfish. 



