PHYSICAL OCEANOGRAPHY OP THE GULP OP MAINE 



689 



However, a cold layer of this same sort, though not so low in temperature, can equally 

 be produced in any partially inclosed boreal sea. All that is requisite is that the 

 surface layers be exposed to a rigorous winter climate, alternating with rapid solar 

 warming in summer, over depths great enough to allow a more or less constant in- 

 flow of warmer ocean water below the level to which winter cooling penetrates 

 (Bigelow, 1917, p. 237). 



In the Baltic, for example, a cold layer reminiscent of the previous winter's 

 chilling persists at a depth of 50 to 100 meters until well into the summer (Knudsen, 

 1909; Kriimmel, 1907, p. 471; Witting, 1906); but increasingly active vertical circu- 

 lation, which accompanies the cooling of the surface after August, entirely dissipates 

 this stratum of low temperature there by late autumn, just as happens in the Gulf 

 of Maine. The following serial temperatures for the Alland Deep (in the Baltic) in 

 winter, spring, summer, and autumn, are introduced for comparison with the 

 Gulf of Maine.^8 



A cold mid layer of the same sort persists into the summer in the Black Sea, 

 where it is self-evident that cold Arctic currents play no part in the temperature cycle 

 and where, consequently, the low temperatures recorded at 60 to 100 meters in 

 August must be purely the product of local influences, as Andrusoff (1893) has 

 pointed out. 



With melting ice no more important in the Black Sea than it is in the Gulf of 

 Maine, ^' the cooling agent chiefly responsible must be the loss of heat from the sur- 

 face by radiation during the cold months. 



The general account of temperature, and especially the temperature sections for 

 the western basin in successive months (fig. 5), makes it clear that the cold layer 

 recorded in summer in the Gulf of Maine reflects the persistence of the low tempera- 

 ture to which the whole upper 100 to 150 meters is chilled in winter, but which is 

 obliterated by autumn, just as happens in the Baltic. No connection appears 

 on the profiles between the development of this cold layer in the western side of the 

 gulf as the spring advances, and the inrush of Nova Scotian water into the eastern 

 side.^° 



«« From Kriimmel (1907, p. 471), after Witting (1906). 



"Tlie northwestern bays and liarbors of the Black Sea (e. g., Odessa Gulf and Kherson Bay) usually freeze over part of the time 

 each winter; but ice very seldom extends more than 2 or 3 miles seaward, and even these shallow areas of low salinity are sometimes 

 open all winter, while the open sea south of the Crimean peninsula never freezes (British Admiralty, 1897). Consequently the 

 amount of ice that actually melts in the Black Sea proper each spring is so small that we can hardly suppose it has any appreciable 

 eflect on sea temperature there. 



'» In an earlier report (Bigelow, 1917) I referred to the Gulf of St. Lawrence as a thermal example of this same sort; but Hunts- 

 man's (1924 and 1925) more recent hydrographic studies indicate a greater inflow of icy water from the Labrador current through 

 the Straits of Belle Isle than Dawson's (1907 and 1913) earlier observations of the strait had suggested. Consequently, the per- 

 sistence into the summer of the minimum layer there, close to 0° in temperature at about 100 meters' depth, results at least in 

 part from the cold water flowing in and from the melting of the Arctic ice which this brings with it in winter and early spiing, as 

 well as from winter chilling and the melting of ice frozen locally within the Gulf of St. Lawrence. 



