PLANKTON OF THE GULF OF MAINE 443 



(station 20109), and in 1913 it was prominent in the rich diatom flora over the north- 

 west part of the banks during the last few days of the month, as noted above (p. 422). 

 In May, 1915, it was not uncommon among the more plentiful Chffltoceras and Tha- 

 lassiosira in the deeps of the gulf (stations 10267 to 10269) and was dominant locally 

 there on the 10th (station 10273) and near the Isles of Shoals on the 14th (station 

 10278). It was also recorded in Ipswich Bay on the 8th in 1920 (station 20122), 

 but it was not detected at all on the western part of Georges Bank and neighboring 

 basin, in the Massachusetts Bay region, in the coastal belt north and east of Cape 

 Elizabeth, nor off western Nova Scotia during that month, either in 1915 or 1920. 



Rh. semispina was not found among the abundant diatom flora of the Mount 

 Desert region in June, 1915 (e. g., station 10285), or in the offshore parts of the gulf 

 during that month, but there was a scattering of it among the Thalassiosira and 

 Chsetoceras in Petit Passage on the 10th, and it might fairly be classed as domi- 

 nant over German Bank on the 19th (station 10290). 



Our midsummer records for this species are confined to Georges Bank (where 

 occasional cells were noted in July, 1914, stations 10219 and 10223, but none at all 

 among the Rh. styliforrnis, Rh. shrubsolei, and Thalassiothrix longissima that swarmed 

 on July 23, 1916); to the Eastern Channel (station 10227), Browns Bank (station 

 10228), the neighborhood of Lurcher Shoal (station 10245), the northeast corner of 

 the gulf (stations 10247 and 10248), the waters off the coast of Maine east of Cape 

 Elizabeth (station 1025S) ; and to the shelf off Marthas Vineyard, where it swarmed 

 on August 25, 1914 (station 10258; fig. 125). Like diatoms generally, Rh. semispina 

 practically vanishes from the central deeps of the gulf during the summer. Nor is 

 there any reason to look for a considerable augmentation in its numbers there during 

 the autumn, for it has appeared only sparingly in our September, October, and 

 November hauls (station 10047, November 20, 1912; stations 10317 and 10336, 

 September 15 and October 26, 1915; and stations 10400 and 10403, November 1 

 and November 8, 1916). It was widely distributed over the northern half of the 

 gulf (always, however, in very small numbers) in the midwinter of 1920-21, when it 

 occurred at about 50 per cent of the stations (stations 10490, 10491, 10494, 10495, 

 10496, 10497, 10500, and 10502). Fritz (1921a) records a scattering of "Rh. hebe- 

 tata," which probably were this variety, at St. Andrews in every month except 

 November. 



The most notable feature of the occurrence of Rh. semispina in the Gulf of Maine, 

 as outlined by our data, is its irregularity; no definite succession of flowerings is 

 demonstrated. On the whole, however, it can be described as at its maximum 

 during the spring and summer (this half of the year includes all the rich flowerings 

 we have encountered), and at its minimum in autumn and winter. At Woods Hole, 

 too, Fish (1925) reports the richest flowerings of this species as occurring in summer. 

 This parallels its seasonal status in northern European seas, where it is most abun- 

 dant from April until June, flowering earliest in the more southern and latest in more 

 northern waters. 84 But no definite correlation between flowering periods and latitude 

 or temperature is yet apparent for the Gulf of Maine. 



•' Flowers most abundantly in tue North Sea in May, but not until August in Greenland waters and in Barents Sea. 



