458 PROCEEDINGS OF THE GEOLOGICAL SOCIETY. 



the common salt encloses some small crystals of the bichromate, like 

 fig. 14. Occasionally these so interfere with the crystallization, that 

 they give rise to small attached fluid-cavities, as seen on the right- 

 hand side of fig. 15 ; or there may be more than one crystal and a 

 larger fluid-cavity, like fig. 16. These fluid-cavities differ from those 

 just described, in containing no small crystals of the bichromate, that 

 which could not be retained in solution having been deposited on the 

 larger preexisting crystals. Moreover, as shown by fig. 16, these en- 

 closed crystals project beyond the general boundary-line of the fluid- 

 cavities, whereas in the other case they are merely deposited on the 

 surface of the cavities. Great numbers of that kind do also occur; 

 and often so many crystals are deposited, that the cavity appears to 

 be quite full of them, as shown by fig. 17. It is, however, well 

 worthy of remark, that many of the smaller cavities remain for months 

 full of a deep yellow liquid, which 1 have supposed to be represented 

 by the shading in fig. 13, containing far more of the bichromate in 

 solution than can be retained when in large quantities, as if these 

 minute cavities exercised the same influence in preventing the deposi- 

 tion of crystals, that, according to Dr. Percy's and my own observa- 

 tions (Dec. 1857), minute tubes exercise in preventing the freezing 

 of water until the temperature is much lower than that at which it 

 freezes at once in larger tubes. 



When, instead of bichromate of potash, chloride of potassium is 

 employed, small cubic or rectangular crystals are deposited in the 

 fluid-cavities of the common salt, as shown in figs. 18 and 19 ; and 

 in every case that I have seen, their edges are all parallel to the rect- 

 angular planes of the cavities. If a concentrated solution of hydro- 

 chlorate of ammonia be used, the fluid-cavities in the common salt 

 contain crystals of the salt of ammonia of a very rounded character, 

 as shown by fig. 20, so as to appear like enclosed globules of a dense 

 liquid. 



These experiments (March 1858) therefore show that, when cry- 

 stals are formed at an elevated temperature, evidence of it is afforded 

 by the contraction of the fluid enclosed in the cavities giving rise to 

 a vacuity, and the reduction of its solvent power causing the deposi- 

 tion of crystals. Since, of course, the amount of the contraction of 

 the fluid depends upon the height of the temperature from which it 

 has cooled, the relative size of the vacuity must indicate how much 

 the temperature at which the crystals were formed was above that at 

 which they are examined, in the same manner as the sinking of the 

 mercury in a self-registering-maximum thermometer shows the dif- 

 ference in the temperature. 



Figs. 21 & 22 in nitrate of potash, 23 in binoxalate of ammonia, 

 24 and 25 in sulphate of zinc, represent fluid-cavities of rather 

 striking forms, — the shading in figs. 24 and 25 being like the ap- 

 pearance produced by planes inclined to the line of vision. In many 

 substances the fluid-cavities are commonly in the form of tubes, 

 which are often of irregular width, and, as it were, pass into rows of 

 smaller, shorter cavities, as shown by fig. 26, representing a portion 

 of alum. Some of the cavities in this salt are as shown in fig. 27; 



