80 WRITINGS OF JAMES SMITHSON. 



ON THE CRYSTALLINE FORM OF ICE. 



From Thomson's Annals of Philosophy, Vol. XXI; Now Series, VoL 

 V, 1823, page 840. 



March 4, 1823. 



Sir: I have just seen a memoir in the Annales de Chi- 

 mie et de Physique for Oct. 1822, but published about a 

 month ago, on the crystalline form of ice. 



Mr. Hericart de Thury is said to have observed ice in 

 hexagonal and triangular prisms ; and Dr. Clarke, of Cam- 

 bridge, in rhomboides of 120° and 60°. 



M. Haiiy supposed the form to be octahedral, and so did 

 Rom6 de I'lsle; and, if I mistake not much, there ia in an 

 ancient volume of the Journal do Physique by Rozier, an 

 account of ice in acute octahedrals. 



Are these accounts and opinions accurate ? 



Hail is always crystals of ice more or less regular. When 

 they are sufficiently so to allow their form to be ascertained, 

 and which is generally the case, it is constantly, as far as I 

 have observed, that of two hexagonal pyramids joined base 

 to base, similar to that of the crystals of oxide of silicium 

 or quartz, and of sulphate of potassium. One of the pyra- 

 mids is truncated, which leads to the idea that ice becomes 

 electrified on a variation of its temperature, like tourma- 

 line, silicate of zinc, &c. 



I do not think that I have measured the inclination of 

 the faces more than once. The two pyramids appeared to 

 form by their junction an angle of about 80 degrees. 



Snow presents in fact the same form as hail, but imper- 

 fect. Its flakes are skeletons of the crystals, having the 

 greatest analogy to certain crystals of alum, white sulphuret 

 of iron, &c., whose faces are wanting, and which consist of 

 edges only. 



In spring and autumn ; that is, between the season of 



