Vol. 6$.'] CONSTANTS IN MINERALS AND IGNEOUS ROCKS. 159 



Moreover, the conditions under which a pegmatite must have 

 consolidated are entirely different from those obtainable in the 

 laboratory. Here, then, the analogy with a saline solution fails. 



In the cooling-down of a magma, the final consolidation of which 

 gives rise to less than 70 per cent, of orthoclase, the quartz ought, 

 according to the eutectic hypothesis, to be the first to solidify, and 

 no orthoclase should crystallize until the composition of the magma 

 has reached eutectic proportions. In the granites, however, the 

 orthoclase is found to have crystallized before the quartz. This 

 must, in great part, be due to the extreme viscosity of the latter and 

 to supercooling of the magma, combined with the better crystallizing 

 properties of orthoclase. Viscosity is, therefore, the great drawback 

 in determining melting-points by the method of direct observation. 



If it were possible to heat the glasses very gradually over a long 

 period of time, the eutectic mixture might be seen to melt more 

 readily than the others. This, owing to the heating of the 

 resistance-wires and consequent diminution of the current, and also 

 owing to the injurious effect which long observations of the ribbon 

 have on the eyesight, is hardly possible with the meldometer. The 

 thermo-electric method, by which Messrs. Day & Shepherd have 

 recently determined the eutectic proportions of the lime-silica 

 series, is, I imagine, more likely to give satisfactory results. 

 Prof. Dcelter (5), however, experimenting on mixtures of minerals, 

 has found that the melting-points of the mixtures are intermediate 

 between the melting-points of the constituents, and concludes that 

 there is no evidence that the eutectic hypothesis can be applied to 

 molten silicates. The same fact is shown by the results obtained 

 with the meldometer, namely : — 



Melting-point of OrQ 6 1317° 0. 



OrQ 3 1292° C. 



OrQ 2 1285° 0. 



.. Or 2 Q 3 1265° 0. 



Or-glass 1220° C. 



On the other hand, a mixture of orthoclase aud albite, in the pro- 

 portions of 40 per cent, orthoclase to 60 per cent, albite (which, from 

 analyses of cryptoperthites, has been determined as their ' eutectic 

 mixture '), when fused into a glass and its melting-point compared 

 with those of albite- and orthoclase-glasses, yielded the following 

 results : — 



Melting-point of albite-glass 1268° 0. 



„ ,, orthoclase-glass 1220° C. 



„ ,, cryptoperthite-glass . 1175° C. 



In this case, then, there seems to be a distinct lowering of the 

 melting-point and a correspondence to the fifth curve of Bakhuis 

 Eoozeboom (8). 



Since the conditions which govern the production of eutectic 

 mixtures in igneous magmas are so little known, it would appear 

 that the construction of apparatus by which a nearer approach to 

 plutonic conditions may be attained must be attempted before any 

 definite result can be achieved experimentally. 



