888 



APPENDIX. 



® )ao:n 



Too to 5*0 °^ an mc h> few being as large as the ^, the breadth being 

 as one to three. 



The various forms and patterns of the snow-flower are due in great 

 measure to the amount of saturation of the air stratum in which they 

 are congealed. The more complete the saturation the larger and 

 coarser the formation ; the smaller the degree of relative humidity the 

 finer and more beautiful the patterns. The so called " diamond 

 dust "falling from a clear sky shows the prism in its most perfect 

 condition : each little amorphous-looking mass is really made up of a 

 number of crystals sometimes showing the most beautiful patterns, to 

 which the large snow-flower, by comparison, is crude and rough. 



The method of attachment between the crystals is usually butt to 

 butt, and they do not often lie across one another. The adhesion be- 

 tween any two crystals of that character is called striction. Very 

 small hexagonal scales under the microscope are seen to be built up of 

 A A' a number of what appear to be smaller 



hexagonal plates, but which are really 

 the ends or butts of a number of hexag- 

 onal prisms which you are looking down upon, the thickness of the 

 scale being the longitudinal diameter of the prisms. 



Frequently you will be able to make out in the crystal (by proper 

 manipulation of light) the vacua marked in the accompanying sketch, 

 •• A and A'." giving the two views. 



I account for the formation of vacua in the following way : When 

 the molecule (as I am inclined to think the molecular group which 

 forms the practical working unit of water) loses heat the process is as 

 follows : — 



Let Fig. I. represent a section through the middle of the molecular 

 group, and which simply for convenience we can conceive to be formed 



of only two layers of molecules, 

 A. and B. Now as the outer 

 layer of molecules A. must ne- 

 cessarily lose heat sooner by rad- 

 iation than the inner layer B.. 

 the temperature of any molecule 

 A. will fall below tbe tempera- 

 ture of any molecule B. that lies 

 within. As soon, however, as 

 this occurs, molecule B. will give 

 n j > some ot the excess ot neat to A., which will again radiate it; and 

 so on, until B. gives up all of its heat, A. receiving the last unit, and 



some of the excess of heat to A. 



