268 



BULLETIN OF THE BUREAU OF FISHERIES 



Cautioning the reader that the difference may be partly exphcable as evidence 

 of "rich" and "poor" years for the species, the percentages indicate that it is prac- 

 tically universal in the inner half of the gulf throughout the summer and early 

 autumn but less plentiful during winter and spring. The average number per 

 square meter likewise shows it to be most abundant in the inner part of the guff 

 during the warm months. 



The average numbers of P. parvus per square meter in vertical hauls, counting 

 only the stations where it occurred, are as follows: 



Date 



March, 1920. 



April, 1920 



May, 1915 and 1920 

 June, 1915 



Average 

 number 



455 



600 



3,656 



1.015 



Date 



August, 1915 



September, 1915 

 October, 1915... 



Average 

 number 



14,042 

 4,065 

 9,045 



If the table were made to include the stations where it was absent, or at least so 

 rare that the vertical net failed to take it, the discrepancy between March and April 

 and the other months would be still greater. The hauls for February, 1920 (sta- 

 tions 20044 to 20048), are omitted from this table because the high average resulting 

 from them (about 2,000 per square meter) is due to catches of 5,000 and 3,000 per 

 square meter at the two stations outside the continental edge (stations 20044 and 

 20045), which would undoubtedly be several times too high for the inner waters of 

 the gulf at this season. 



In the western side of the basin Paracalanus increased in number in 1915 from 

 about 1,000 per square meter on May 5 (station 10267) and 1,300 on June 26 (station 

 10299) to 16,100 on August 31 (station 10307). 



In the eastern side of the basin where there were only about 1,100 Paracalanus on 

 June 19 (station 10288) the vertical haul took 23,450 per square meter on August 6 

 (station 10304). On September 29 there were 850 per square meter at a station in 

 Massachusetts Bay (10320), and the number had risen to about 14,000 by October 

 27 (mean of stations 10338 and 10339). A change of the opposite order at a neigh- 

 boring location near Gloucester, where the number per square meter declined from 

 more than 25,000 on May 4 (station 10266) to about 2,500 on August 31 (station 

 10306) and about 3,000 on October 1 (station 10324), shows how the formation and 

 dispersal of local shoals may more than offset the general seasonal augmentation of 

 the species at any particular locality. 



Off the Isles of Shoals a slight decrease took place from 5,250 per square meter on 

 May 14 (station 10278) to 3,170 on October 4 (station 10325); on German Bank the 

 figure remained about stationary from May 7 (1,500 per square meter at station 10271) 

 to June 19 (1,500 at station 10290) and September 2 (1,600 at station 10311). 



Notwithstanding these irregularities, not one of the October stations yielded less 

 than 2,000 P. parvus per square meter, and the maxima within the gulf were much 

 greater in October (30,750 off Cape Cod, station 10336, and 24,450 in Massachusetts 

 Bay, station 10338) than in September (6,650 per square meter, station 10319). 

 Thus it seems that there are actually more P. parvus in the gulf in mid-autumn than 



