OF ARTS AND SCIENCES : FEBRUARY 11, 1873. ' 511 



tributed upon the ice, as in the case of a roadway across a stream or 

 pond, the transit may be safe along the beaten track long after horses 

 and vehicles would sink on either side. The ice in the latter case had 

 been sheltered from the sun. 



Areas of ice swept of a light fall of snow will become soft at times 

 when neighboring ice, protected by the snow, will retain its glare and 

 hardness. The significance of the fact of exposure to the sun's rays 

 will presently appear. 



Some sixteen years ago I began the study of this problem by taking 

 a series of photographs of the vertical surfaces of blocks of ice, trans- 

 parent and apparently thoroughly solid, or nearly so, throughout, as 

 they slowly melted under the direct rays of the sun. These photo- 

 graphs, successively taken at intervals of about half an hour, showed a . 

 constant approach to columnar structure. 



I exposed water to be frozen in vessels of various shapes, and photo- 

 graphed the ice obtained, and found the columnar structure in general 

 perpendicular to the outer surface of the ice. "When the surface was 

 convex below, the columns curved upward toward the centre of the 

 mass. 



While pursuing this mode of investigation, I conceived the idea of 

 taking plaster casts of the decaying ice. By placing the blocks of 

 honeycombed ice in boxes and pouring liquid plaster, such as is used 

 by stucco-workers, over the ice, the plaster flowed in to take the place 

 the melted ice had left, and " set" In a few hours the enclosed ice 

 melted, and the water flowed out or escaped by evaporation. On per- 

 mitting the cast carefully to dry, which was facilitated by removing the 

 upper crust of plaster and exposing the interior to the air, and then 

 touching the slender rods of plaster with a camel's-hair pencil dipped 

 in a weak solution of gum-arabic, and repeating this process for several 

 times with alternate drying, the whole cast at length acquired such 

 consistency as to admit of being moved about with safety. The forms 

 which were thus obtained from ice in various stages of decay, from . 

 large open ponds and from pools, were of great beauty and full of 

 material for study. 



I desire chiefly in this note to present what I conceive to be the so- 

 lution of the problem of this columnar structure, that others who may 

 perhaps have better opportunities than myself may apply the explana- 

 tion and test it. 



When water, by cooling at the surface, has throughout obtained its 



