SCIENTIFIC RESULTS 125 



measurements of the temperature may be made to give warning of 

 the proximity of bergs. An electrical resistance thermometer grad- 

 uated to one-thousandth of a degree centigrade called a microthermo- 

 gram was experimented with on a trip to Hudson Bay in July, 1910. 

 Barnes (1910, p. 131) shows that a typical thermograph upon ap- 

 proaching a berg begins with a slight rise and then drops from 1.0° C. 

 at 1 mile, to 2.2'^ C. at one-half mile. The primary peak and then the 

 fall in temperature were explained as due to Pettersson's No. (1) 

 current, i. e., to the thaw water spreading out on the surface, the outer 

 fringe being heated most by the sun, was warmest. But Barnes's 

 (1913) further investigations from the Canadian Pacific steamship 

 MmitcaJm near the Strait of Belle Isle the summer of 1912 found, 

 contrary to his earlier experiments, a continuous rise of temperature 

 as bergs were approached suggesting the drop in the thermographs 

 noted in 1910 was actually due to the generally low temperature of 

 the cold Labrador current in which the bergs floated. Barnes con- 

 cludes therefore that Pettersson's current Xo. (1) does not prevail 

 under actual conditions at sea, but instead that there is an infloAving 

 surface current toward a berg. Aitken (1913, p. 561) carried out a 

 laboratory experiment similar to Pettersson's, finding that fresh 

 ice melting in salt water induces a circulation whereby all of the thaw 

 comes to the surface. Taylor (1914. p. 65). employing a microther- 

 mogram similar to Barnes's, failed to secure any definite reliable tem- 

 perature graphs when the Scotia approached close to icebergs on the 

 Grand Banks in the spring of 1913. Sometimes there was a rise, 

 sometimes a drop, and usually the record was masked by fluctuations 

 in the temperature due to causes other than icebergs. Thuras (1915, 

 p. 67) approached bergs on the ice patrol ship using a thermometer 

 graduated to one-tenth of a degree centigrade, finding that no definite 

 rise and fall of the instrument was re])eated with regularity. Cruis- 

 ing past some bergs the instrument would remain quite steady, while 

 at others it would be very erratic. Aitken (1915, p. 561) criticizes 

 Barnes's work in vieAv of the contradictory results obtained by him- 

 self and by Taylor. 



Barnes (1927, p. 92), in a recent article describes the circulation in- 

 duced by melting icebergs as follows : 



Every icebers is a hydraulic pump sinking the surrounding sea water by cool- 

 ing and drawing to itself the warm surface waters and thereby contributes to 

 its own destruction. The warmed surface layers flow more rapidly to the cold 

 ice surface than do the cool layers, hence the iceberg becomes the central point 

 for the collection of the warmer surface water. 



By his view the warmed surrounding water is an indraft from the 

 surface layers which compensates for the sinking thaw water. This 

 horizontal movement keeps them uppermost and explains their 

 greater insolation and warmth near a melting iceberg. 



Ricketts (1930, p. 119) discusses the conventional theory of circida- 

 tion around icebergs and he finds it difficult in view of the stratifica- 

 tion and marked stability of the surface layers to believe that melting 

 icebergs establish currents of any appreciable magnitude. He as- 

 sumes, however, for the sake of argument, that an average-sized 

 berg south of Xewfoundland chills a 10-foot layer of the surrounding- 

 water to a temperature of 20° F. He then shows that under the given 

 conditions the resulting horizontal indraft will be 27 feet per day 



