APPLICATIONS OF CRYSTALLOGRAPHY, ETC. 571 



as indicated above little absolute difference of kind exists 

 between liquids and gases, and it is consequently advisable, 

 before attempting to apply laws governing these two 

 states to solid substances, to inquire whether the differences 

 between solids and liquids are also, in the main, differences 

 of degree rather than of kind, and. if considerable differences 

 of kind do exist, to ascertain whether they are of a nature 

 likely to vitiate arguments based on analogy. 



At the very outset of the inquiry we are met by a well- 

 marked difference of kind between solids and liquids. The 

 particles composing a liquid are free to move, and under 

 ordinary conditions bear no fixed and definite orientation 

 one to the other ; althouo-h Perkin's work on magnetic 

 rotation probably indicates that under certain conditions, 

 as when placed in a magnetic field, the particles of a liquid 

 do arrange themselves in some regular manner. Almost 

 all solid substances, however, are crystalline and an amor- 

 phous or non-crystalline solid is a substance of some rarity, 

 most substances which are casually described as amorphous 

 being in reality crystalline. That the arrangement of the 

 particles of a crystalline body is not a structureless or 

 amorphous one is evident from a very superficial examina- 

 tion of such properties of crystals as the colours, solubilities, 

 electrical properties, cleavages and refractive indices ; these 

 properties being different in different directions in the 

 crystal demonstrate conclusively that the arrangement of 

 the parts or structural units of the crystal varies with the 

 direction. And we are now, thanks to the labours of 

 Sohncke, Barlow and others (compare Miers, Science 

 Progress, 1894, i-- 4^3 ; 1896, iii., 129), acquainted with all 

 the possible kinds of arrangement of these units of the crys- 

 talline structure. It is important to consider in what way this 

 homogeneous arrangement existing in crystalline substances 

 but not in liquids or gases can affect analogies drawn be- 

 tween liquids and gases on the one hand and crystalline 

 bodies on the other ; the best method of showing the kind 

 of disturbance introduced and the ways by which it may 

 be eliminated is probably to consider a few actual cases. 



The careful study of the refractive indices of liquids and 



