NOTE ON THE PRINCIPLE OF REGELATION, ETC. 69 
tion of that body, already taken as a type of simple glass; while 
granular sugar, and more perfectly sugar-candy, exhibit its crystal- 
line state. Prof. Graham has shown that, at a certain temperature, 
by mechanical means the former may be converted into the latter, 
the temperature quickly rising 70° on the transition of the sugar 
from the glassy to the crystalline state. This and similar facts in- 
duced him to refer the peculiar constitution and properties of glass 
in general to the permanent retention of acertain quantity of heatin 
a latent state, which becomes sensible on its crystallization ; and this 
will take place on its being preserved in a soft state at certain 
temperatures. 
There are some remarkable and instructive parallels between the 
phenomena of the crystallization of water, and that of glass and 
some other bodies. It follows from the experiments and inductions 
of Gregory Watt already cited, that during the crystallization of 
glass a higher temperature must be communicated to the interior 
than that existing over its surface, by the evolution of heat at the 
points where the crystalline form is assumed, which will be gradually 
conducted throughout the mass. So that, in the express words of 
Faraday, in relation to ice, ‘‘ by virtue of the solidifying [erystal- 
lizing] power at points of contact, the same mass may be freezing and 
thawing at the same moment ;” and the “freezing process in the 
inside may be a thawing process on the outside,” and thus contribute 
to the slowness of the cooling, and allow the crystallization therefore 
to be the more perfect. We here seem to have the explanation 
of the well-known fact, that in bodies which crystallize from a state 
of igneous fusion, the most perfect crystalline state is produced 
when the longest time intervenes between the commencement of 
solidification (now using that term in its ordinary sense) and the 
complete cooling of the melted mass. The cases cited from Mr. 
Faraday at the beginning of this paper, of the growth of crystals 
(including those of ice in ice-cold water) in solutions, all have their 
exact parallels in the accretion of crystals in cooling melted glass. 
“ Crystals of ice,” Mr. Faraday observes, “ which could not be colder 
than, the surrounding fluid, exhibited the phenomena of regelation.” 
—that is, of incorporation into one—“ when purposely brought in 
contact with each other.” The same thing happens with melted 
glass slowly cooling, in which crystalline spherules, often forming 
spontaneously and independently, continue to form and to increase, 
