PLANKTON OP THE GULF OF MAINE 
201 
summer, consequent on fluctuations in its actual abundance, and in the abundance 
of the other species of copepods. Twenty-one vertical hauls at as many stations 
for September and October, 1915, give an average of only 38 per cent and 42 per 
cent of Calanus, respectively — that is, little more than half of the May percentage. 
The percentage of Calanus averaged somewhat higher in the horizontal hauls 
of December, 1920 (about 58 per cent; table, p. 304). However, this does not reflect 
an increase in the actual abundance of the species (which, on the contrary, decreases 
markedly in numbers during the late autumn and early winter), but a still more 
pronounced decrease in the local stock of other species of copepods. Thus, while 
curves for the actual and for the relative abundance (percentage) of C. -finmarchicus 
would be similar for the spring, they would be contradictory for the September- 
December quarter, and to this extent the percentages taken by themselves would 
give a totally false picture of the seasonal fluctuations of the species in the Gulf of 
Maine. 
From the economic standpoint this means that any copepod-eating fish in the 
Gulf of Maine is likely to make Calanus its chief diet from May until August and 
in October, but to dependless on it and more on othercopepods during the early autumn 
and again in late winter and early spring. 
The average percentages need further qualification to bring out the great 
irregularity in the relative abundance of the species which we have encountered 
from station to station on most of the cruises and from month to month at individual 
stations, irregularities connected with the streaky way in which 0. finmarchicus 
often occurs, and with the formation and dissipation of its shoals. In Massachusetts 
Bay, for example, the percentage fell from 80 to 45 at one locality off Gloucester 
between March 1 and April 9, 1920, but increased from 6 per cent to 50 per cent 
off Boston Harbor, near by, during approximately the same interval. In the western 
basin, at three successive stations, the percentage of C. finmarchicus was 90 on 
February 23, 25 on March 24, 75 on April 18, and at three stations along a line run- 
ning out from Ipswich Bay toward Platts Bank, on April 9 and 10, the percentages 
were alternately 75, 25, and 80. Seventy-five per cent of Calanus in the south- 
western part of the basin on February 23, but only 2 per cent on the neighboring 
part of Georges Bank the same day, was evidence of a corresponding difference in 
the actual number of C. finmarchicus per square meter — respectively, 6,562 and 
25 — but on the southeastern slope of the bank the percentage fell only from an 
average of about 75 per cent on March 12 to 60 per cent on April 15, although this 
interval saw the dissipation of a very dense swarm of Calanus, occasioning a shrinking 
in the number per square meter from 103,000 to about 600. 
Apart from the question of vertical stratification (p. 24), the percentages of 
Calanus have proved more nearly uniform over considerable areas in the later 
spring and summer. In early May, 1920, for example, it constituted 60 to 80 per 
cent of the copepods at most of the stations in the southwestern part of the gulf 
and on the western end of Georges Bank (table, p. 302). In July and August, 1914, 
its percentages in the horizontal hauls at most of the stations inside the continental 
edge approximated the average (71 per cent) for all the stations, irrespective of 
regional variations in the actual abundance of the species. In September and the 
