460 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



are precisely like those formed below the level of the solution, whilst 

 others are as if it had been partially evaporated in the cavities before 

 they were finally enclosed in solid crystal ; and therefore they contain 

 much solid matter, as shown by fig. 37. Minute crystals of sulphate 

 of lime are also enclosed in the solid crystal ; and some have been 

 deposited inside the fluid-cavities before they were finally shut up, 

 as shown in fig. 36. Other cavities have caught up air as well as 

 fluid, so that they contain very large bubbles, as shown in fig. 34 ; 

 whilst in others, like fig. 35, there is no fluid ; — these various kinds 

 gradually passing into one another. In a similar manner when the 

 solution of salt contains less bichromate of potash than can be held 

 in solution at the ordinary temperature, there occur, in the crystalline 

 crust, cavities like those in crystals formed below the level of the 

 liquid when a dilute solution is employed, as well as like those when 

 a saturated hot solution is used, on account of its becoming variously 

 concentrated by drying on the surface of the crust. 



b. Number, size, form, and arrangement of cavities. 



There is generally a most intimate relation between the number of 

 cavities in a crystal and the rate at which it was formed. This is 

 well illustrated by the chlorides of sodium and potassium ; for when 

 very slowly deposited, they are transparent, and contain but few, 

 whereas, when deposited more quickly, they are so full of cavities as 

 to be very white and opake. In some cases the deposition proceeds 

 rapidly at first, and white opake nuclei are formed ; and afterwards 

 it proceeds more slowly, and the exterior of the crystals is clear and 

 transparent, as shown by fig. 1, — the change from opake to trans- 

 parent being either sudden or gradual, according to circumstances. 

 This also usually happens when substances are crystallized by the 

 cooling of a strong hot solution ; for then deposition proceeds 

 rapidly at first, but slowly towards the close of the process. Sul- 

 phate of potash, however, contains nearly as many cavities when 

 formed slowly as when deposited quickly. There is also a consider- 

 able difference in the number of cavities in different salts, though 

 formed under similar conditions. Thus, if solutions of alum and 

 chloride of potassium be evaporated moderately quickly side by side 

 at the ordinary temperature, the chloride of potassium contains so 

 many cavities as to be perfectly white and opake, whereas the 

 alum contains very few, and is perfectly transparent. The same 

 is the case when a mixed solution of common salt and alum is 

 evaporated. 



In general the size of the cavities varies inversely as their number ; 

 for when the crystals are slowly formed, they are larger, though less 

 numerous, as shown to great advantage by the different parts of 

 fig. 1. If the rate of growth be the same, crystals formed at a high 

 temperature contain fewer and larger cavities than those formed at a 

 lower. 



When the sides of the cavities are definite and straight, they are 

 planes of the crystal ; and therefore there is a connexion between the 



