PLANKTON OF THE GULF OF MAINE 415 



taking place a month or so earlier in the season. With similar emendation in date in 

 one direction or the other, it would apply to Massachusetts Bay equally in an "early" 

 year, such as 1913, or in a late year. 



The mutual fluctuations of C. tripos and C. longipes may be summarized as 

 follows for the Gulf of Maine as a whole: 



Early in spring when the vernal augmentation of diatoms is at its height, Cera- 

 tium (and indeed all peridinians) practically vanishes from the gidf, an event taking 

 place lirst along the northwest coast, where the diatoms llower earliest, and soon 

 afterwards in other parts of the gulf. After the flowerings of diatoms dwindle, C. 

 tipes (fig. 112) multiplies until July, when all the gulf, except for a narrow zone 

 along its northeast and east coasts, supports an abundant Ccratium plankton. 

 During July C. tripos (figs. 113 and 114) multiplies in the central deeps. As the 

 summer advances the area of abundance of C. tripos expands coastwise and the stock 

 of ft longipes dwindles until tripos becomes predominant along the southwestern, the 

 eastern, and finally along the northwestern and northern coasts of the gulf, with ft 

 longipes persisting latest as an important factor in the plankton in the region between 

 Cape Elizabeth and the Grand Manan Channel. C. tripos predominates throughout 

 the winter, but even then, when C. longipes is at its lowest ebb, the latter has occurred 

 in small numbers at most of our stations; nor does either species vanish wholly 

 from the gulf at any season, though either may be so scarce when the other is at 

 its peak of abundance, as well as during the flowering period of the diatoms, that 

 careful search of considerable amounts of plankton may be required to reveal its 

 presence. 



The seasonal changes in the relative abundance of these two peridinians must 

 not, of course, be understood to take place in as orderly a manner as they are repre- 

 sented here, for they are undoubtedly accompanied by temporary interruptions and 

 even reversals, which would alter the smooth curves to a succession of zig-zags, were 

 daily or weekly records available. In fact, such a reversal is known to have taken 

 place in 1915 oif Machias, Me., where longipes was predominant on July 15 (in the 

 proportion of 16 longipes to 3 tripos at station 10301), was outnumbered by tripos 

 on September 11 (station 10316), was again predominant on October 9 (station 

 10327), and would doubtless have been found outnumbered by tripos a month after 

 that, had we visited that region again later in the season. Sporadic alternations of 

 this sort do not weaken the general thesis that the succession, as here outlined, is a 

 regular and characteristic feature of the planktonic cycle of the gulf, however, though 

 its time table varies from year to year, as do all other seasonal changes in the sea. 



In the foregoing account I have purposely refrained from alluding to the status 

 of the two leading species of Ceratium on Georges Bank in late summer or autumn 

 (longipes predominates therein spring and early summer (p. 40S) as it does elsewhere 

 in the gulf), because no collection of phytoplankton has yet been made on the bank 

 during the half year, August to February. 



A fourth species of Ceratium — ft fusus — has been taken so often in our tow 

 nets that it deserves brief mention, though it is never predominant in the Gulf of 

 Maine. C. fusus has been found at most of the stations where the genus as a whole 

 75808—26 27 



