1815.) On Crystallization. 103 
more than an hour; yet by then ¢urning the vessel, so as to pour 
out a little from the neck, the crystallization has immediately oc- 
curred. 
9. The same solution in the mattrass above mentioned has fre- 
quently become completely crystallized when left uncorked; at 
other times a large mass, equal to half the volume of the solution, 
has crystallized regularly, in hard transparent crystals, the remainder 
of the solution continuing fluid. 
10. Saturated mixed solutions of nitre and Glauber’s salt, corked 
closely, have allowed the nitre to crystallize regularly at the bottom; 
whilst the Glauber’s salt remained fluid, and on drawing the cork 
became solid in the usual way. 
11. Solutions, by no means saturated, evince similar results with 
the above fully saturated ones, although not in so strongly marked a 
manner. 
12. One of the most singular and interesting facts connected with 
these experiments is, that in those cases in which (either in the 
corked or uncorked solutions) regular, firm, transparent crystals 
form, so soon as the residuary saturated solution above them solidi- 
fies, either spontaneously, or by shaking, drawing the cork, &c. an. 
immediate (or nearly so) opalescence, or loss of transparency, ensues 
in those first formed crystals, which gradually increases to a beau~ 
tiful porcelainous whiteness. This I have almost invariably noticed 
under the above circumstances: J believe it arises from the gradual 
abstraction of the water of crystallization of the first formed regular 
crystals, by the mass of secondary crystals; for in one experiment 
made, I found the porcelainous mass, when dissolved in water, and 
regularly recrystallized, afforded a quantity of transparent crystals, 
superior in weight to those I employed, which could only arise from 
their re-obtaining their thus lost water of crystallization. How the 
secondary crystals operate in withdrawing this water from the first, 
I cannot form the most distant idea. 
13. In those solutions in which spontaneous crystals have formed, 
in the course of a few days, if the secondary crystallization does not 
take place, a complete truncation of the summits of the crystals 
occurs, gradually forming a level of the whole, as in common 
eases ; yet in several instances the solution above was sufficiently 
saturated to consolidate when shaken, 
14. In ove experiment two equal sized phials were filled to the 
top with saturated solutions; one was corked, the other was left 
open: in two hours the uncorked one had consolidated; the other 
was observed to have contracted above one-fourth of an inch, and 
continued fluid; it crystallized, however, as usual, when briskly 
shaken, without withdrawing the cork. 
It should perhaps be mentioned, that this sudden crystallization 
always commences at the surface. ; 
_Thave put the solutions, both corked and uncorked, into cold 
water, as soon as made, in, order to expedite their cooling, and 
have found the same results generally as when suffered to cool gra- 
