454 BULLETIN OF THE BUREAU OF FISHERIES 



though dominant and extremely plentiful in April, it practically vanishes by the 

 first week of May. 



Neither of the two most abundant species of Thalassiosira {Th. nordenskioldi or 

 Th . gravida) exists planktonic in the surface waters of the open gulf in any numbers 

 after August, nor are they recorded for the outer parts of the Bay of Fundy after 

 August 10 by Bailey (1917). The fact that we found a scattering of Th. norden- 

 shioldi close to Swan's Island off the mouth of Penobscot Bay on September 15 

 (station 10317), and again in Massachusetts Bay on October 1 (station 10322), 

 during the autumn of 1915, shows that they may persist in small numbers here and 

 there along the coast until well into the autumn; but the genus has not been detected 

 in any haul in any part of the open gulf between the last week of October and the 

 first week of February. 70 



The seasonal lists of Thalassiosira in northern seas generally (especially the 

 well-marked periodicity in its appearances and disappearances), and the certainty 

 that its abundance in the Gulf of Maine results from local flowering and not from 

 immigration, makes it probable that is passes the balance of the year, from the close 

 of the summer flowerings until its reappearance in the plankton in early spring, on 

 the bottom as resting spores. But so far as I am aware these have not actually 

 been seen in this genus. 



The relative numerical proportions in which the two commoner species of Thalas- 

 siosira — Th. nordenskioldi and Th. gravida — occur in our spring and summer samples 

 have not been worked out fully, but the preliminary examination suggests that on 

 the whole Th. nordenskioldi is the more important in March, April, and May (as 

 might be expected from the experience of European students), and that Th. gravida 

 increases in relative abundance as the season advances. A third species of Thalas- 

 siosira, Th. decipiens, which has been rare in the spring tow nettings (stations 20059, 

 20101, and 20104 to 20106), appeared in numbers near Mount Desert Island (station 

 10328) and off Penobscot Bay (station 10329) on October 9, 1915. Th. hyalina has 

 been detected at several widely separated localities during the spring of 1920 (occa- 

 sional specimens only) — viz., off Cape Cod on March 24 (station 20088), in the 

 Northern Channel (station 20105), over Browns Bank on April 16 (station 20106), 

 off the southeast face of Georges Bank on April 16 (station 20109), and in the Eastern 

 Channel (station 20107). Thalassiosira baltica is recorded from one station (20061) 

 and may well have been overlooked elsewhere among the swarms of Th. nordenskioldi. 

 There is also one locality record each for Th. clevei (station 10328), Th. svbtiles 

 (station 20089) , and Th. Uoculata n (station 20107) . 



Thalassiothrix 



This genus is represented in the Guff of Maine hauls by two species — longissima 

 and nitschioides. The records for Th. longissima are too scattered to outline its 

 seasonal fluctuations in our waters in a satisfactory way. It appeared only twice in 

 the catches for March, 1920 — that is, in the southeast corner of the basin (station 

 20064) and on Georges Bank (station 20066). McMurrich (1917), too, found it 

 only once at St. Andrews during that month (March 4) ; then, however, in abundance. 

 It was not detected among the Thalassiosira- Chsetoceras flowerings in the north - 



; ° Fritz (1921) records it on Feb. 9. ;1 Identified by Dr. Albert Mann. 



