PHYSICAL OOEANOGKAPHY OF THE GULP OF MAINE 653 



These several midwinter stations (fig. 80), combined, show that at this season 

 any line run normal to the coast of the gulf would lead from lower surface temper- 

 atures out into slightly warmer water, with the surface then coldest (below 1°), 

 locally, close in to the land between Boston and Cape Elizabeth on the one side of 

 the gulf, and along Nova Scotia on the other; slightly warmer than 4° along the 

 intervening coast sector, outside the outer islands, and about 6° on the central and 

 southern parts of the basin (fig. 80) ; but the temperature may fall as low as 1° 

 among the islands by the end of December, as happened at Boothbay and in Lubec 

 Channel in 1919 (figs. 30 and 31). 



These local differences result from the topography of the coast line, from the 

 local winter climate, and from differences in the activity of vertical stirring by the 

 tides. Thus, the surface chiUs more rapidly at the head of Massachusetts Bay than 

 along the open coast of Maine because less actively mixed by the tides with warmer 

 water from offshore and from deeper levels. Chilling takes place most rapidly of all 

 in the sounds and harbors, because their enclosure prevents free interchange with 

 the water outside. 



In midwinter the surface is, as a whole, the coldest level, though differing by 

 less than 1° from the warmest stratum at most of the stations. Thus, the inner 

 part of Massachusetts Bay (station 10488) had cooled to 3.89° at the surface on 

 December 29, with 5.86° on the bottom in 60 meters. In the bowl off Gloucester 

 the readings were 5.56° at the surface and 6.9° to 7° from 40 meters down to the 

 bottom in 150 meters, the latter almost precisely reproducing the temperature recorded 

 there on December 23, 1912 (fig. 75). The surface was about 0.5° warmer 15 

 miles off the northern end of Cape Cod (station 10491), but the 100-meter level was 

 about 0.1° cooler. The vertical distribution of temperature was the same near the 

 land, off the mouth of the Merrimac Kiver (station 10492), as near the head of Mas- 

 sachusetts Bay, and with the actual values nearly alike, while the trough off the Isles 

 of Shoals (station 10493, fig. 70) agreed equally with the sink station off Gloucester 

 just mentioned. 



The vertical range of temperature was only about 0.2° off Seguin in about 80 

 meters depth on December 31, 1920 (station 10495, 5.83° on the surface, 6.1° at 

 40 meters, and 6.1° at 75 meters); but a few mUes farther out from the influence 

 of the land off the mouth of Penobscot Bay, the next day (station 10496), where 

 the water is less subject to tidal stirring, the temperature curve closely paralled that 

 for the Isle of Shoals station 2 days previous in the upper 100 meters (5.6° at the 

 surface, 6.05° at 40 meters, and 6.79° at 100 meters), but showed a slight vertical 

 warming at greater depths to 7.5° on the bottom in 150 meters. Surface (4.7°) and 

 90-meter readings (5.7°) differed by about this same amount close in to Mount 

 Desert Island (station 10497). However, the temperature was uniform, surface to 

 bottom, a few miles off Machias (station 10498, 5.56° to 5.61°), a state approxi- 

 mated here throughout the year. 



In the Fundy deep the Halcyon found the whole column about 1° to 2° warmer 

 on January 4, 1921 (station 10499), than Mavor (1923) records it for January 3, 

 1916; in fact, agreeing more closely with his temperatures for December 5, 1918, in 

 spite of the difference in date, as follows: 



