183 
Proceedings of the Roycil Irish Acadony. 
hare proved this point, at a given pressure, is entirely dependent as 
regards this group (and therefore as regards M. Debray's experiments 
with ferric oxide) upon the relative volume of water. 
So sensitive is the potassio-ferric alum, that a fine amethyst coloured 
crystal, obtained by crystallizing it at a low temperature, when placed 
upon the palm of the hand, first becomes white, and after a short time 
yellow. This salt is, of course, the most sensitive of all the group 
mentioned in this paper. 
It may be well to point out that these experiments can only 
determine the laws of dissociation, practically, to a limited extent. 
"We must build upon them, and conceive in our mind's eye the 
enormous results that are attainable with the aid of unlimited time 
and pressure. We are compelled to bring into play the Scientific 
use of the Imagination," and bridge over the long gap that lies 
between our mimic results and those consummated in the laboratory 
of nature, and by this means to mentally extend our experience. The 
results are here, but the rnodus operandi is absolutely a sealed book. 
I propose to reserve the consideration of the element silicon more 
particularly in connexion with the bases dealt with in this paper. 
Addendum A. 
On the Transition of Compounds from the Colloid to the Crystalloid 
Condition. 
In Part I. of this report reference has been frequently made to the 
fact, that ferric, chromic, and aluminic oxides, and their compounds, 
are capable of taking the colloidal form. 
E'ow, there is a peculiar character of crystallization discerned in 
minerals, which is always indicative that the constituents had been 
previously in their colloidal condition, and had afterwards become crys- 
talloidal. In most cases this change proceeds from assimilation of 
water of crystallization, but not necessarily so. The characteristic 
which I wish to specify is where the crystals radiate from a centre, 
and form in the aggregate what may be termed the radial spherical 
system. It includes many of the forms technically known as nodular, 
mammellated, botryoidal, and reniformed masses. Such minerals are 
formed, in most cases, from aqueous solutions, which had evidently 
been at one period under the dissociating influence of heat. The iron, 
manganese, and lime minerals, and also the zeolites exhibits these pe- 
culiar forms most strikingly ; but none more so than that beautiful 
mineral, wavellite. The mode by which this form of crystallization is 
produced cannot be more strikingly exhibited than by an organic salt, 
which I described some years ago, viz., chlorate of quinia. Although 
this solution is organic, the formation can be more easily observed in 
Buch a case than in the minerals, where time and pressure are such 
