136 



E. C. CASE 



There seems but one explanation of this very queer phenom- 

 enon: the balls were rolled by the waves. The beach is a flat, 

 smooth sand beach, and the water is very shallow for a considerable 

 distance out. It seems that in some snowstorm early in the winter, 

 before any ice-foot had formed, the beach must have been covered 

 with 2 or 3 inches of a snow (the layers of the snowballs are from 

 an inch to an inch and a half in their present compacted condition). 



Fig. 2 — A large snowball about 3 feet in diameter, half melted and showing con- 

 centric structure. 



in a very damp and soggy state, perhaps filling the water adjacent 

 to the shore with a heavy slush. The water must have been very 

 close to the freezing-point, so that there was little or no melting 

 of the snow as it came in contact with the water. Now, a rise of 

 the wind would produce a surge which, moving up the beach and 

 back, started the snow in motion, and as the snow was water-soaked, 

 and too heavy to float with any buoyancy, it was pushed back and 

 forth until it was compressed into a small mass which began to 

 roll. There is no distinct nucleus to the balls either of harder 

 snow or of small pieces of ice, as might be expected, but the center 



