of Crystallized Bodies. 287 



in a parallel position, where longitudinal delineations appear- 

 ed, whereas they were six-sided in the spaces, formed by the 

 meeting of them, and wherever the delineation appeared exe- 

 cuted on a smaller scale. I have attempted to express some- 

 thing of this kind in Fig. 22. The ice had much more accu- 

 mulated, where the six-sided' plates were visible,- so as con- 

 siderably to obstruct the passage of light. There were even 

 crystals, like the one marked a, with a distinct geniculated 

 appearance. 



The explanation of these phenomena can be scarcely 

 given but upon the supposition of prismatic crystals. There 

 might exist two species, characterized by different forms, with 

 the same constituents of hydrogen and oxygen, as we have 

 analogous cases in calcareous spar and arragonite, or in 

 the hexahedral and prismatic iron-pyrites ; but so imperfect 

 is our positive information in this respect, that we are not 

 entitled to consider such a hypothesis as plausible. There 

 is one hypothesis, indeed, which might suffice for explain- 

 ing all the appearances hitherto observed. The forms of 

 ice might belong to the pyramidal system. The individuals 

 forming on the surface of stagnant water would have all their 

 axes parallel, and consequently show the single system of rings 

 in polarized light equally well as if they did constitute one 

 single individual. For the six-sided prisms, or six-sided plates, 

 as produced by regular composition, and the star-like figures 

 of snow, we find many analogous cases in the pyramidal sys- 

 tem, for instance in tin ore and in rutile. The latter particular- 

 ly often shows compressed prisms, and reticulated aggrega- 

 tions disposed upon a plane. Actual observations, however, 

 and measurements of angles, are required to show whether this 

 opinion be borne out by fact. Rome de ITsle,* Bosc d , Antic,-f- 

 and Scoresby,J have each observed four-sided pyramids, which 

 would add to the probability of the forms of ice belonging to 

 the pyramidal system. 



(To be continued.) 



* CrystuUographie, t. i. p. 4. 



+ Journal de Physique, 1788, t. xxxiii. p. 57. 



$ Mem. Wern. Soc. vol. ii. part 2(1. 



