438 BULLETIN OF THE BUREAU OF FISHERIES 



plentiful and more generally distributed in the coastal belt east of Penobscot Bay 

 in autumn than we have found it in August. McMurrich and Fritz have likewise 

 found it comparatively plentiful at St. Andrews during the last half of October; 

 in fact, Fritz's counts locate its plurimum for the year at that season. 



We have no evidence that this autumnal augmentation of Coscinodiscus extends 

 to the western part of the gulf, our October and November stations west and south 

 of Penobscot Bay having yielded few or none during the seasons of 1912, 1915, and 

 1916. Its duration must be short even in the St. Andrews region, also, for both 

 McMurrich (1917, p. 9) and Fritz (1921) found it considerably less plentiful there 

 in November than in October, but it must multiply in early winter, being widespread 

 from late December on. 



The several species of Coscinodiscus are so closely allied to one another that the 

 determination of them must await future critical study, wrong identifications being 

 worse than none. The reader will find above (p. 423) lists of those so far determined 

 by Doctor Mann for representative stations and seasons. 



Coscinosira 



Coscinosira polychorda, a neritic species, has occurred sparingly among the 

 Thalassiosira and Chagtoceras at the April and May stations in both sides of the 

 gulf (stations 20090, 20093, 20095, 20096, 20103, 20104, 20106, and 20107 in April, 

 1920, and 10277 on May 14, 1915) and at one June station (102S5, June 14, 1915), 

 always near land. We have never found it an important factor in the spring phyto- 

 plankton, but it was relatively abundant, if not dominant, off Swan Island near 

 Mount Desert Island on September 15, 1915 (station 10317), and occasional speci- 

 mens were also noted at the same general region on the 9th of the following month 

 (station 1032S). One well-preserved chain was also noted in a haul off Machias, 

 Me., January 4, 1921 (station 10498). The only Georges Bank record for Cos- 

 cinosira is for April 15, 1913 (Bigelow, 1914a, p. 415). Bailey (1915, pi. 2, fig. 15; 

 pi. 3, fig. 4) figures it from the Bay of Fundy, and Fritz (1921) includes it under the 

 general heading "Thalassiosira" in her lists of diatoms for St. Andrews. 



Ditylium 



The genus Ditylium is never more than a minor factor in the plankton of the 

 open Gulf of Maine, but it deserves a brief word here because it is an excellent indi- 

 cator of waters of coastwise origin, being strictly neritic, but at the same time able 

 to survive long sea journeys thanks to its powers of flotation, and so easily recognized 

 that it is not apt to be confused with any other diatom. 



The Gulf of Maine records for it are confined to the immediate vicinity of the 

 western and northern coasts, mostly inshore from the 100-meter contour (fig. 127). 

 Ditylium is not known either from Nova Scotian waters on the east or from the 

 offshore banks on the south. 



As a rule, the records of Ditylium outside the outer islands have been based on 

 occasional specimens only among more plentiful diatoms of other genera. It was 

 comparatively abundant in Massachusetts Bay on March 4, 1921 (station 10505), 

 and Fritz (1921) found it plentiful at St. Andrews in October, with a scattering in 

 November. 



