SCIENTIFIC EESULTS 41 



the bathynu'ti'ical features of tlie shelves are important factors in the 

 track of tliis stream of })ack ice. BafHn Bay. a shaUow e!on<i;atecl 

 basin coA^ering Got) mih's of latitude, connected Avith the Arctic Ocean 

 by the narrow openin<>- between Greenland and Ellesmere Land, and 

 in a less direct way throu<»:li the maze of Arctic sounds to the west- 

 ward, is one of the chief reservoirs of the ice. Davis Strait is the 

 bottle neck throug^h which a o-reat }>roportion of the pack emerjzes 

 into the Xorth .Vtlantic. Anotlier ])i-<)bable souice is the rej^ion of 

 Hudson Strait and Fox Channel. Fhe Labrador current is the 

 jireat aa'eucy of trans[)oi-tation southward alon^- Labrador, past 

 Xewfoundland, and over the Grand Bank. It is the low temperature 

 of the Avater over the continental shelf from Baffin Land to the 

 Grand Bank, approximately 150,000 square miles, the surface layers 

 of wdiich, are chilled to 0° C. (32° F.) or lower, that permit the 

 southward drift of the pack. The extent of these frigid shelf waters 

 and of those of Bafhn Bay furnishes a clew to the aggregate annual 

 output of sea ice. If two-thirds of the total 467.000 square miles of 

 the ice area is nornuilly covered to a depth of feet, the eastern 

 North American pack consists of ai)i)roximately 450 to 475 cubic 

 miles of ice yearly. (See fig. 121. p. 200.) 



The several tributary sounds located on the western side of Baffin 

 Bay and Davis Strait contribute relatively great quantities of pack 

 to the eastern Xorth American ice stream. During the colder 

 months these openings supply -' ice to the southward moving 

 masses, but in summer their discharges create areas of 0})en water. 

 These channels from north to south are as follows : Smith Sound, 

 Jones Sound. Lancaster Sound, Hudson Strait, and Strait of Belle 

 Isle, The sounds of Baffin Bay contribute the greatest quantities of 

 ice. while the straits to the south are res])()nsible for the greatest 

 amount of dissipation and Avastage. Summermeltingisusually vicAved 

 as a ])henomenon spi'eading from south to north. I^ut in case of the 

 Xorth American pack, account must be taken of the disrui)tive influ- 

 ences along its Avestern flank. Thus the dischai'ge from Hudson 

 Strait A^ery much hastens the dismemberment of the pack in that 

 offing. The Marion expedition in 192S had an excellent ojjportunity 

 to survey the results of these disintegrations. As early as June 11 the 

 (rodthaah expedition arrived and found the Avaters open off Hudson 

 Strait and the ice lying quite far back both to the north and to the 

 south. Sometime shoi-tly ])i'ior to the (lodthnah's A'isit, the strong 

 currents ])ouring in and out of the strait had ai)i)arently severed 

 the ice stream and isolated the large Labradoi- Held. On August IS, 

 about tAvo months later, the Marion exi)edition found the offing of 

 the strait still clear; and the Labrador field had disappeared by that 

 time, though the southern edge of the Baffin Land pack remained 

 about the same as it had been a month earlier. 



In 102s the ])ack around C'a])e Dier. Baffin Land, consisting of 

 IS.OOO scjuare miles of ice. never ])enetrated farthei" south than Oum- 

 berland Gulf after June, showing that the rate of its soutlnvard 

 adA'ance Avas offset by the rate of melting in the sun-heated waters 

 discharged through Hudson Strait. The isolated Labrador held of 



=" MeckinK (1906. pp. 22-31 and appended map) after consulting the accounts of the 

 various early explorations of the Amtrican Archipelago has constructed a map showing 

 the prevailing circulation. Mecking shows that practically all of the currents flow, and 

 therefore the ice masses move, from west to east; that is, from the polar regions into 

 Baffin Bay and Davis Strait. 



