C.—-GEOLOGY. Th 
and alumina; fireclay contains more alumina with a small and variable 
amount of alkalis. Portland cement is a mixture of silica, alumina, and 
lime. All these manufactures are produced by pure dry fusion under 
atmospheric pressure, and exactly under the conditions and tempera- 
tures of the experiments on which the diagram is founded. The 
presence of small amounts of impurities in the natural minerals em- 
ployed introduces complications, but these may be neglected if only 
approximate results are aimed at. During the War the highly trained 
technical skill of the workers of the Geophysical Institute and their 
refined apparatus were entirely at the service of American industries, 
such as the manufacture of optical and chemical glass, and all the 
practical problems that arose were promptly and satisfactorily solved. 
The investigation of the system Ca-Al,0,-Si0, which we have taken 
as an example of the best type of modern work in this field of research. 
is a great contribution to theoretical petrology. It has bearings on the 
thermal alteration of many rocks such as quartzites, flints, pure lime- 
stones, siliceous and argillaceous limestones, bauxites,  fireclays, 
calcareous quartzites, all of which may be regarded as mixtures of 
silica, clay, and (carbonate of) lime together with their alteration pro- 
ducts. But for the geologist as a rule the matter is not quite so simple, 
and caution is necessary in drawing inferences. Three of the 
commonest alteration products in this group of rocks, for example, are 
biotite, garnet, and andalusite, and these minerals have been produced 
experimentally only under very exceptional conditions. 
There are differences between the conditions of the experiments and 
those that actually obtain in the making of rocks, and these are 
essentially of three kinds: 
(a) Experimental work is successful only when the conditions are 
exactly defined, and necessity compels us at present to restrict experi- 
mental work to simple systems of two or three components. The theory 
of these has been very fully worked out, and this is essential to the 
interpretation of the experimental results. Systems of four components 
have hardly yet been touched. If we take, for example, the four 
common oxides of rocks, CaO, Al,O:, MgO, and SiO,, the six possible 
binary systems are pretty well known, and the four possible ternary 
systems have also been thoroughly studied, but little progress has yet 
been made with the investigation of quaternary mixtures containing all 
four components. The mathematics of such a system is of the most 
complex description. Now the common rocks contain seven or more 
components, and their behaviour in igneous melts is a problem which is 
at present beyond solution. 
To simplify matters we might investigate such a system piecemeal, 
that is to say, we might take parts of it and treat them as independent 
systems. For example, the three minerals anorthite, forsterite, quartz. 
which consist of these four components, have been investigated, and 
it was proved that this could not be regarded as a simple three-com- 
ponent system, as under certain conditions phenomena appeared which 
characterised a quaternary mixture. Along these lines, however, there 
is no doubt that great progress can be made, and the results already 
obtained are so valuable that they hold out great promise for the future. 
