758 ABSTRACTS OF PAPERS 



dition in 1899. Work on the geology and ore deposits of ttiis district for the 

 United States Geological Survey has given opportunity for further study of 

 these glaciers and the mapping of the fronts of all of them. As a rule there 

 has been since the growth of the present forest a period of retreat in the case 

 of these glaciers. One, the Barry, has its front 2 miles farther back than in 

 1899. The best known of these glaciers, the Columbia, was advancing rapidly 

 in the summer of 1909 into the forested zone along its front. 



OSCILLATIONS OF ALASKAN GLACIERS 

 BY B. S. TARE AND LAWRENCE MARTIN 



(Abstract) 



The National Geographic Society's Alaskan Expedition of 1909 observed the 

 following glacial oscillations. In Yakutat Bay the Marvine lobe of Malaspina 

 glacier and the Atrevida, Haenke, and Variegated glaciers have ceased the 

 advance which began in the winter of 1905-1906. The Hidden glacier has ad- 

 vanced over 2 miles since 1906, but has now begun to shrink away from the 

 new shore moraines. Hubbard glacier has advanced slightly. Lucia glacier is 

 newly crevassed and advancing this summer (1909) and is riding up on a 

 nunatak. These oscillations confirm the earthquake-avalanche theory for 

 glacial advance, proposed in 1906 by the senior author, and furnish facts as to 

 the brevity of such advances. On the lower Copper River the Childs glacier 

 was more active in 1909 than 1908, but the position of the front remains un- 

 changed. The Miles, Childs, and Baird glaciers are essentially as in 1884. 

 1885, 1891, and 1900. In eastern Prince William Sound the Valdez and Shoup 

 glaciers are slowly receding. The Columbia glacier has advanced rapidly since 

 1908 and is building moraines and destroying the forest, as was observed by 

 Prof. U. S. Grant early in 1909 and by the National Geographic Society expe 

 dition later in the season. The events in the glaciation of Prince William 

 Sound differ decidedly from those in the Yakutat Bay region. 



Discussion 



Dr. F. E. Matthes : The cause assigned for the spasmodic advance of the 

 Alaskan glaciers by Professors Tarr and Martin is no doubt a competent one 

 and probably the true one; at the same time it is not a necessary one. The 

 mode of advance of glaciers is normally spasmodic. Why this should be is 

 not clear, as the mechanics of glacial flow involved is as yet imperfectly under- 

 stood. The fact, however, has been satisfactorily established by experiments 

 on ice-flow under regulated pressure recently made by German investigators 

 and described by Professor Hess in his treatise "Die Gletscher." These experi- 

 ments show that ice confined in a reservoir remains inert under steadily in- 

 creasing pressure until a certain point is reached when flow sets in, slowly at 

 first, but increasing rapidly in velocity, even though the pressure remains con- 

 stant or is diminished. The flow then continues with gradually diminishing 

 velocity until the reservoir is well depleted, when the ice-mass returns to its 

 inert state. The conditions in nature very closely parallel those obtaining in 

 the experiments. The annual overflow from a glacial cirque does not corre- 

 spond to the annual accretions; but the snow keeps on accumulating for 



