252 



BULLETIN OF THE BUREAU OF FISHERIES 



several months previous. Probably the specimens in question had drifted thither 

 around Cape Cod from the center of abundance in the southwestern part of the gulf. 



Granting that M. Tonga is able to breed in the gulf to some extent, its periodic 

 disappearances are sufficient evidence that it does so only sporadically and tempo- 

 rarily. Perhaps it is only able to carry on through one or two generations in the 

 high temperatures in which it must exist there, and failing accessions of new stock 

 dies out until there is a fresh invasion from the north. Evidently such fluctua- 

 tions in local reproduction and migrations mirror the physical features of the water 

 in which this little crustacean lives, but it is not yet possible to state the precise 

 relationship which its temporary appearances in the Gulf of Maine bear to tempera- 

 ture and salinity there or in the waters to the east and north, or to the seasonal or 

 annual variations in the flow of the currents. 



There is every reason to class it a cold-water species in the gulf, and it has 

 actually been taken there in water a fraction cooler than zero (at St. Andrews, 

 February, 1917; Willey, 1921); but having been found widespread in the summer 

 and autumn of 1915 in temperatures as high as 8 to 10°, it can survive and perhaps 

 even breed over a wider range than has generally been supposed in European seas, 

 where 6.75° is the highest temperature of record for it (Farran, 1910), and where 

 most of the captures have been from water of 2.25 to 3.25°. M. longa was in 

 comparative abundance and apparently in good condition off Marthas Vineyard at 

 14.5° (station 10331), but it is hardly conceivable that it could have lived long there. 



Minimum temperatures at any depth at stations where Melridia longa is recorded for August, September, 



or October, 1915 



More information is needed before the relationship between the salinity of 

 the water and the occurrence of M. longa in the gulf can be traced. Most of the 

 records for this species in the northeastern Atlantic have been from salinities rather 

 higher than those of the Gulf of Maine, where it has been taken most commonly 

 in water of 32 to 33.5 per mille; but Nordenskiold's account (p. 248) suggests that in 

 the very low temperatures of the polar sea it may be able to exist in water but slightly 

 saline, and we took it in salinities of 31 to 32 per mille on several occasions during 

 the spring of 1920 and once in 29.94 per mille (station 20096, surface haul). Probably 

 M. longa is never plentiful enough to be of much importance in the natural economy 

 of the Gulf of Maine, but no doubt it serves to some extent as fish food, having been 



