420 BULLETIN OF THE BUREAU OF FISHERIES 



that Coscinodiscus was most numerous over the inner half of the shelf where it, 

 together with the oceanic species Chsetoceras criophilum and Ch. decipiens, composed 

 the bulk of the catch (stations 20073 to 20075), but occurred only sparsely at the 

 outer stations (stations 20076 and 20077) ; whereas neritic species (notably Ch. mitra, 

 Ch. diadema, Ch. laciniosum,, Ch. debile, and Thalassiothrix nitscMoides) were most 

 plentiful over the outer half of the shelf (stations 10275 and 10276), not next the 

 land as one might have expected, and even occurred outside the continental slope 

 as well (station 20077). 



Such a concentration of neritic forms at the outer stations off Shelburne instead 

 of at the inner is intelligible when hydrographic conditions are taken into account, 

 because the axis of the cold Nova Scotian current of low salinity, itself essentially 

 neritic in its biologic aspect, occupied precisely the same location at the time. 



The abundant diatom community already mentioned (p. 383) as characterizing 

 the western part of Georges Bank on February 23, 1920, consisted chiefly of slimy 

 masses of the tiny neritic species Chsetoceras sociale, not of Coscinodiscus nor of the 

 oceanic species of Chsetoceras, though Ch. decipiens, Ch. criophilum, and Ch. atlan- 

 ticum all occurred there, as did the neritic forms Rhizosolenia shrubsolei, Eucampia 

 zodiacus, and Leptocylindrus. This flowering of Ch. sociale was very local, as seems 

 usually to bo the case when concentrations of diatoms occur on Georges Bank, and 

 was confined strictly to the comparatively shoal waters of the bank (stations 20046 

 and 20047). Ch. sociale was sought in vain in the tow netting over the edge only 

 20-odd miles distant (station 20045), where Thalassiothrix nitschioides and an 

 occasional cell of Guinardia and Chsetoceras diadema were the only neritic diatoms 

 recognized. The very sparse community of diatoms in the basin immediately to 

 the north of the bank (station 20048) consisted of the same oceanic species of diatoms 

 that characterize the central parts of the gulf generally in February and March — that 

 is, Coscinodiscus, Chsetoceras atlanticum, Ch. criophilmn, Ch. decipiens, Ch. boreale, 

 Ch. densum, Rhizosolenia semispina, and Thalassiothrix longissima. 



No tropical phytoplankton was found at our stations outside the continental 

 slope in February or March, 1920 (stations 20044, 20069, and 20077). 



Our work for 1913 had already suggested that the diatoms that first commence 

 rapid multiplication in the Cape Ann-Cape Elizabeth region in spring are the fore- 

 runners of the vernal flowerings that are the most spectacular event in the yearly 

 planktonic cycle of the Gulf of Maine. These are the several species of Chsetoceras 

 that may rival the peridinians here and there along the coast even as early as the last 

 of January or early February, especially in Ipswich Bay. Shortly thereafter the 

 genus Thalassiosira begins flowering, a phenomenon which we have been able to 

 follow through parts of the years 1913, 1915, and 1920. 



In 1920 the tow at the mouth of Massachusetts Bay contained Thalassiosira, 

 besides several other kinds of diatoms, on March 1 (station 20050; see list p. 423); 

 and Thalassiosira and Chsetoceras must both have commenced flowering actively 

 even earlier than this alongshore between Cape Ann and Cape Elizabeth that year, 

 the "rich" diatom area outlined on the chart (fig. 104) being dominated by these 

 two genera on March 4 and 5. 



The list given below (p. 425) for the station near Cape Elizabeth (20059), which 

 was paralleled near the Isles of Shoals (station 20060), and the dominance by Thalas- 



