TiCHBORNE — On Dissociation by Heat of Compounds. 185 
tion in solutions of any considerable strength on boiling ; but if 
a very considerable proportion of water is there to the amount of salt, 
a basic compound thereof is precipitated at 100° C. This is best seen 
by passing a beam of electric light through the flask, as the precipitate 
formed is almost optically invisible in daylight. A beam of electric 
light impinging upon the light flocculent precipitate at once reveals its 
existence. It is almost absolutely necessary to use some such light 
when operating upon chloride of aluminium ; but an aluminic alum, 
by virtue of the extra play of affinities brought to bear, gives a reaction 
easy of observation by the eye. These precipitates redissolve on cool- 
ing. As regards the effects of extraordinary heat procurable under 
pressure, we find that alumina behaves in a similar manner to the 
other members of this group, which the following experiment will 
illustrate: — 
Experiment XX. : — 
The ammonia and potash alums were placed in separate sealed 
tubes, and submitted for two and a half hours to a temperature of 
177° C. White crystalline precipitates were produced, varying in 
composition, but containing sulphuric acid. They represented the 
greater part of the alumina present. 
JSxperiment XXI. : — 
An open tube containing a solution of chloride of aluminium was 
placed in the same apparatus (described in the iron experiments), and 
submitted for some hours to a temperature of 177° C. At the line of 
juncture of fluid and steam, a white colloidal substance was deposited, 
which seems to consist of a mixture of hydrated and anhydrous alu- 
mina. This experiment throws some light upon the formation of 
corundum. Sapphire, ruby, and topaz, have been formed artificially 
with perfect success by Ebelmen, Deville, and others. All these 
artificial processes, however, have been igneous ones. In nature, 
however, it is evident that the corundum is not always, if it is ever, 
formed in this manner. It frequently contains from 3 to 4 per cent, 
of water, and is also associated with diaspore, and other hydrated 
minerals. 
Before concluding this part of my report, I will consider for 
a moment the results of these experiments collectively. It would seem 
that the compound molecules of aluminic, chromic, and ferric oxides, 
are dissociatable in ratio to their atomic weight, or we may say, as regards 
this group, that their basylous position is in ratio to their dissociatability. 
Thus, although compound alumina molecules are dissociatable at ordinary 
temperatures, when insufficiently dilute, the decomposition is not so 
well marked until we get an extraordinary pressure, and a correspond- 
ing increase in the range of temperature. Again, that in the presence 
of other compounds of basylous elements they are more easily dis- 
sociated from the introduction of a second basylous molecule reacting 
upon the acid. The following series of experiments upon the alums 
and simple salts will place what is stated in a clearer light. Some of 
them were performed by taking very diluted but standard solutions of 
R. I. A. mOC. — TOL. I., SER. II., SCIENCE. 2 B 
