58 BULLETIN OF THE BUREAU OF FISHERIES 



copepods (p. 56), never more than two tropical animal species among the hosts of 

 boreal animals. 



This scarcity of planktonic visitors of the tropical category within the Gulf of 

 Maine and even over its shallow southern rim, when so rich a tropical surface fauna 

 inhabits the inner edge of the Gulf Stream along the outer edge of the continental 

 slope only a few miles without the 100-fathom contour, is fundamentally due to their 

 inability to survive or to reproduce in the low temperatures of the coast water. 

 Their sporadic and solitary occurrence there, contrasted with the considerable 

 numbers and even communities of tropical planktonic animals that often drift close 

 inshore west of Cape Cod, is explicable only on the assumption that the surface 

 waters of the Gulf Stream very seldom overflow the barrier formed by Georges Bank, 

 an assumption corroborated by the physical character of the water. Nevertheless, 

 the Gulf of Maine does owe to the tropical water indirectly, if not directly, one 

 common and very characteristic summer visitor, the large chretognath Sagitta serrato 

 dentata. This species, which is the dominant member of its systematic group in the 

 coastal waters south of New York, occupies a rather peculiar faunal niche in the 

 Gulf of Maine, for while it breeds only in the high temperatures of the Gulf Stream 

 (so far as the area under discussion is concerned), great numbers drift into the cooler 

 mixture zone along the edge of the continental shelf, where they thrive and grow 

 to a much larger size than they do in the warmer waters farther offshore, either 

 because lower salinities and temperatures especially favor their growth (though not 

 their reproduction), or perhaps because of a richer food supply (p. 323, and Hunts- 

 man, 1919). As a denizen of this mixed water, S. serratodentata is swept in abundance 

 into the Gulf of Maine, where, because of its size and abundance, it is the most 

 prominent of all the exotic immigrants, though it never attains a more permanent 

 status there. 



Owing to its peculiar relationship to oceanic temperatures, all the Gulf of Maine 

 records so far obtained for S. serratodentata have been for large specimens, the locali- 

 ties of capture indicating considerable longevity for it within the gulf. It is strictly 

 seasonal in its presence there, however, being so rare in winter and early spring that 

 we have taken it only twice between December 1 and May 1, viz, in Massachusetts 

 Bay on December 4, 1912 (station 10048), and again on January 16, 1913 (station 

 10050). It appears in the eastern side of the gulf as early as the first week in May 

 (p. 320, and Bigelow, 1917, p. 296), and by June it has spread generally over the 

 eastern basin and into the Bay of Fundy as well as over the outer edge of the shelf 

 off Cape Sable, and probably also all along the southern and eastern parts of Georges 

 Bank, where we found it in July, 1914. This species penetrates the inner parts 

 of the gulf so slowly during the early summer that in five years we have found it 

 only once in the western and southwestern parts prior to August 1. Thereafter, 

 however, it spreads so rapidly westward and southward along the coast of Maine 

 that our August and September records for it cover the whole northern half of the 

 gulf from Cape Ann right across to Cape Sable, including Massachusetts Bay, where 

 it occurs regularly in late summer and autumn. 



The locations of the stations of capture and the fact that S. serratodentata is 

 usually more numerous in the eastern than in the western side of the gulf (p. 322) are 



