350 
ON THE FORMATION AND GROWTH OF CRYSTALS. 
soon enable us to define the most suitable condition in this respect. 
In cases where the substance to be crystallized has to be elimi- 
nated, as the crystallizable alkaloids, vegetable acids, &c, the 
density or specific gravity becomes the only test by which to as- 
certain the proper degree of concentration. But here, again, we 
do not always succeed ; for, although we may have found that 
very fine crystals were obtained on an occasion when the solution 
had an ascertained density, yet, unless the apartment into which it 
was set aside shall always have pretty nearly the same tempera- 
ture, the crystals will not be satisfactorily produced. 
So that, having acertained, in the one case, the best proportion 
of saline substance for solution in water, supposing that to be the 
menstruum, and the most suitable density in the other, care must be 
taken that the place where both are set aside, shall have a tem- 
perature subject only to a trifling variation. Underground places 
answer very well, in many instances ; but in others, it is better 
that the temperature be a little higher, since it is found that if 
they be crystallized at a temperature lower than the average, they 
are apt, at a higher one, to give up a portion of their water, 
to become fluid, or to crack. 
These conditions being determined, and a number of crystals ob- 
tained, we have now to consider the best mode of insuring their 
subsequent symmetrical growth. This may be effected in three 
ways, either by adding, at stated intervals, additional quantities 
of the salt dissolved in a small quantity of the solution itself; by 
adding to the solution a fluid capable of uniting with the men- 
struum, but in which the salt is insoluble ; or, by an appropriate 
evaporation, i. e., by reducing the solution to a given weight, and 
returning the crystals when it has sufficiently cooled, or by a con- 
tinuously spontaneous evaporation, in a warm atmosphere, as in 
the cases of protonitrate (nitrate of the sub-oxide) of mercury, and 
sulphate of manganese. 
Next in importance are the arrangements which affect the po- 
sition of the crystals. For prisms (right or oblique) a flat-bottom- 
ed, smooth, shallow dish will be found the most suitable ; for oc- 
tahedrons, a cylindrical vessel, into which is suspended a piece of 
horse-hair or gut (to be had at the fishing-tackle warehouses, and 
known in the trade as " silk-worm gut.") 
Matters being thus arranged, the solution is set aside for (say) 
