APPENDIX. 305 



of effect. All that I observed double might be classed among 

 the higher order of crystals, and I do not recollect an in- 

 stance of a double crystal being closely allied to the primitive 

 stellar figure. 



" It is remarkable that on the morning of the thaw the crys- 

 tals were those originally of a high order, but attenuated and 

 reduced to the faintest possible semblance of, I have no doubt, 

 their original very complex structure ; and this is more than a 

 vague inference, as they continued to collapse and simplify 

 themselves without ceasing until the moment of their disap- 

 pearance, and which evidently was a continuation of the ope- 

 rations which had been constantly in action during their de- 

 scent. The temperature at which I observed them was some 

 degrees above freezing, and the thermometer had only the 

 same morning turned to rise above the low temperatures which 

 had for so many weeks prevailed. The air was at this time 

 genial and mild, more so than might have been expected from 

 the reading of the thermometer. After a week's intermission 

 the frost set in again, but with less rigour, and snow crystals 

 were abundant ; but their entire character was changed, and, 

 with a temperature many degrees below that of the morning 

 of the thaw, there fell, unaccompanied by snow, a large num- 

 ber of primitive stellar crystals, as different in character to 

 those before alluded to, as it is well possible to imagine, for 

 the probable reason that on the morning of the thaw the 

 upper regions of the air were intensely cold, a supposition 

 borne out by the recurrence of the frost after a short interval ; 

 whilst on the mornings when the last- mentioned stellar 

 crystals made their appearance, it is reasonable to suppose 

 that the intense cold above had mitigated ; a conjecture borne 

 out by the decreasing severity of the weather, which, though 

 for many successive weeks considerably below the average for 

 the season, yet soon attained too high a temperature for the 

 continuance of these investigations. The great difference of 

 character between the crystals of the last day of the frost and 

 its recommencement, a week later, is very significant of an 

 entire change of atmospheric conditions. The study of these 



VOL. TJ X 



