426 Da BREWSTER on the Refractive Powers and other 



as those in minerals. I have sometimes observed frozen drops 

 of dew, containing a portion of water which remained unfrozen 

 even at low temperatures ; and I have recently had occasion to 

 examine some crystallisations of ice, which presented the same 

 fact, under more curious circumstances. 



A very sharp frost occurred in Roxburghshire on the morning 

 of the 8th October 1825. The gravel- walks in the garden were 

 raised up about an inch above their natural level by the sudden 

 congelation of the water in the earth mixed with the gravel. All 

 the elevated portions consisted of vertical prismatic crystals of 

 ice of six-sided prisms, with summits which seemed to be tri- 

 edral. The leaves of plants, &c. were covered with granular crys- 

 tals, which were in general six-sided tables. 



Upon examining with a microscope the prismatic crystals ag- 

 gregated in parallel directions, they presented some curious phe- 

 nomena. They had numerous cavities of the most minute kind, 

 arranged in rows parallel to the axis of the crystals, and at such 

 equal distances as to resemble a series of mathematically equi- 

 distant points. Some of the cavities were very long and flat, and 

 sometimes they were amorphous ; but in general they contained 

 water and air. 



Upon submitting one of these cavities to a powerful micro- 

 scope, it appeared as shewn in Fig. 13. of Plate XIX, where 

 A B C is the piece of ice, having in it a long cavity m o, contain- 

 ing water and air. The ice gradually dissolved ; and when the 

 end n o of the cavity m n was near the edge of the ice C B, the 

 air, in a portion of it n o, detached itself, and went off at p, 

 through the solid ice, the cavity closing up again at n. This 

 phenomenon is analogous to the passage of the expansible fluid 

 through topaz and quartz, which has been already described; the 

 air in the one case, and the fluid in the other, finding its way in 

 the direction of easiest cleavage, and the fissure closing up again 

 in the manner already mentioned in a preceding part of this 





