Alumina with Silica, Lime and Magnesia. 295 



results thus far obtained into formulae which were deduced 

 for wholly different conditions and which apply none too 

 accurately even then. Obviously, the calculation of molecular 

 weights from concentrations of ten, twenty, and even fifty 

 per cent, as has sometimes been done, can serve no useful 

 purpose. 



A glance at the technique of measurement of conductivity 

 in aqueous solutions will reveal how unwise are generalizations 

 based on experiments with the conductivity of silicates. Main- 

 taining a constant temperature throughout even a relatively 

 small volume is extremely difficult at a temperature of 1200° C. 

 Electrodes and containing vessels can not be maintained 

 constant in shape or dimensions, nor can perfect or constant 

 contact relations between the electrodes and the melt be 

 assumed. Until such essential conditions can be supplied, we 

 cannot hope to derive much useful data from merely passing the 

 electric current through a silicate. The effect of high and 

 varying viscosity on conductivity is unknown. 



Pyrometry has made great progress in the last decade, and 

 for temperatures below 1600° C. the thermoelement is capable 

 of reading accurately to one-tenth of one degree. But it 

 does not follow that all thermal phenomena in silicates are 

 definable with this precision. The phenomena of melting 

 and inversion, for example, seldom occur with sharpness 

 enough to allow of their being determined within less than 

 il° for compounds, and d=2° for eutectics. We must 

 guard against the too common error of assuming that we have 

 determined the phenomena with the accuracy with which we 

 can read the scale of the instrument. 



Above 1600° C. the optical pyrometer must be used, and the 

 accuracy is much less. With the thermoelement the evolution 

 or absorption of heat which occurs in a charge registers itself, 

 leaving nothing to the judgment of the observer. With the 

 optical pyrometer, as ordinarily used, the melting temperature 

 must be inferred from the apparent fusion of the charge, and 

 the observer is dependent upon his arbitrary judgment as to 

 what constitutes fusion. Usually he cannot determine the 

 beginning of fusion nor can he tell the exact point where the 

 change is complete. Here again, not enough is known of the 

 viscosity of the various mixtures to allow for its effect upon the 

 phenomenon which the observer sees. As Day and Allen have 

 shown, * albite retains its rigidity long after fusion (deorientation 

 of crystal structure) is complete. Quartz acts in a similar 

 manner. The different concentrations in mixtures show all 

 gradations between this extreme viscosity and the extreme 



* Isomorphism and Thermal Properties of the Feldspars, Publications of 

 the Carnegie Institution of Washington, No. 31. 



