462 BULLETIN OF THE BUREAU OF FISHERIES 



few or none even on the same day. The August swarm just mentioned was so con- 

 centrated that only odd specimens appeared in the tow at the station next to the 

 south (station 10256) and none at all at those to the east or north. On September 

 1, 1915, when it was abundant at station 10309, none were taken 40 miles to the 

 southwest (station 10308), 35 miles to the east (station 10310), or 60 miles to the 

 northeast (station 10315). Similarly, none were taken off Cape Elizabeth on Sep- 

 tember 20, 1915 (station 10319), nor off Mount Desert Island on the 15th (station 

 10317), though it was plentiful at an intervening station (10318) on the 16th, with 

 no notable hydrographic difference in the state of the water. There is no apparent 

 correlation between the presence or absence of Acanthometron in the gulf and the 

 precise temperature, for while the August swarms of 1914 were living in water of 

 about 18 to 20° off Cape Ann, the Nova Scotian collections for September, 1915, 

 were from a temperature colder — and perhaps very much colder — than 15°. We 

 have one record of Acanthometron from water of only about 4° (German Bank, 

 station 20103, April 15, 1920). 



Its occurrence is equally independent of salinity within broad limits, for it was 

 most abundant in the Western Basin and off southern Nova Scotia when the water 

 was near its freshest for the year, but we have not detected it in Massachusetts 

 Bay until salinity has increased considerably from its seasonal minimum. Broadly 

 speaking, however, Acanthometron is plentiful in the gulf only while the tempera- 

 ture is comparatively high and the salinity comparatively low. 



Acanthometron likewise attains its seasonal maximum in late summer and 

 autumn off north European coasts, with a general increase from August on, and its 

 minimum in May. On both sides of the Atlantic the richest catches of this radio- 

 larian have been from the eddies of cyclonic currents — that is, from the southern 

 Norwegian Sea (May), Irminger Sea (July), middle of the North Sea (November), 

 and in our gulf from the Western Basin (August) . 



Although Acanthometron occurs at times and locally in even greater abundance 

 in the shallow coastal waters of the eastern side of the Atlantic than in the Gulf of 

 Maine, it is essentially an immigrant there from the open ocean. The records of 

 the International Committee for the Exploration of the Sea prove that although it is 

 independent of actual temperature and salinity for its ability to exist, its abundance 

 in the North Sea depends more or less on the amount cf warm, highly saline ocean 

 water entering around the north of Scotland (Mielk, 1913). Its chief centers of 

 abundance in the Gulf of Maine have been in the regions farthest removed from 

 such oceanic influence — that is, close in to land and in the semistagnant Western 

 Basin. Furthermore, we have never found it in the zone of mixture between cool 

 coast waters and warmer ocean waters along the continental slope, and its absence 

 there is particularly significant because Acanthrometron centers have often been 

 encountered in the contact zone between Atlantic and Arctic waters on the other side 

 of the North Atlantic. 



Here we must leave the question of the distribution of Acanthrometron for the 

 present; but in passing I may point out that more data on this point are particu- 

 larly desirable, not so much for the sake of mere completeness of local information as 

 because of a very interesting phenomenon exhibited by this form, namely, the sharply 



