ON COLLOID CHEMISTRY AND ITS INDUSTRIAL APPLICATIONS. 101 
indefinite colloid, which undergoes further hydrolysis in contact with 
an excess of water, progressively losing lime, until a residue of gelatinous 
silica remains. Dicalcium silicate is rendered more soluble by the presence 
of the soluble aluminate. Tricalcium silicate hydrolyses fairly rapidly, 
the product being of the same general character as in the last-mentioned 
case. A strong mass is obtained, and there is little doubt that pure 
tricalcium silicate would form a satisfactory Portland cement, as was 
in fact observed in the early experiments of Newberry. 
According to the experiments of the Bureau of Standards, then, theinitial 
set of Portland cement is due to the hydration of tricalcium aluminate, 
the product being amorphous. At the same time any sulpho-aluminate 
which may be present crystallises, and any excess of free lime becomes 
hydrated. Within twenty-four hours, the hydration of tricalcium silicate 
begins, usually becoming complete within seven days. Between seven 
and twenty-eight days, the hydration of the dicalcium silicate, which 
is present in considerable proportion in most cements, and the passage 
of the aluminate to the crystalline condition, are the chief stages in the 
process. Of these, the increase in strength is attributed to the formation 
of colloidal material, whilst the change from the colloidal to the crystalline 
state is regarded as involving an actual diminution of strength, the net 
effect being a gain. These results stand in conflict with many previous 
observations, but they have not yet been directly challenged. 
Lastly, reference should be made to the influence of salts and other 
soluble substances on the rate of setting of cements. The addition of 
calcium sulphate to Portland cement for the purpose of retarding the 
setting has long been practised on the large scale, and the action of 
a number of ‘ catalysts,’ both positive and negative, was studied by 
Rohland.1® Michaélis regarded the action as dependent on the change 
of solubility of the aluminates and silicates due to the presence of 
foreign salts, but Rohland,in more recent papers, has attributed it rather 
to their influence on the rate of coagulation of colloids.1® The action ot 
colloidal additions, such as tannin, straw infusion, &c., is very irregular, 
and no connection between the accelerating or retarding influence 
and the chemical composition has been found.2® The changes in setting 
time which many cements undergo on storage, even out of contact 
with moisture, are probably of the same character, and the subject calls 
for much closer investigation than it has yet received. 
NITRO-CELLULOSE EXPLOSIVES FROM THE STANDPOINT OF 
COLLOIDAL CHEMISTRY. 
By E. R. Curystaxz, B.Sc. (Lond.), F.I.C. 
General Review. 
The work which has been done on this subject from a scientific point 
of view is meagre in the extreme. The text-books on colloids barely 
mention the fact that nitro-cellulose forms colloidal solutions with organic 
solvents. The papers published for the most part deal with isolated 
cases investigated for the purpose of elucidating troubles in manufacture. 
'* P. Rohland, Der Portland-Zement vom physikalisch-chemischen Standpunkte, 
Loipaig, 1903. 
* Kolloid Zritschr. 1911, 8, 251; 9, 21. 
* Hi, K, Bonson, C. A. Newhall, and B. Tremper, J. Ind. Eng. Chem., 1914, 6, 795. 
