CunnincHamM—Crystallization of Minerals in Igneous Rocks. 401 
a salt and its solvent can be inverted by varying the relative pro- 
portions of the two substances. Several workers have obtained 
various eutectic mixtures by using proper proportions. Since 
Bunsen’s time this side of the subject has received particular 
attention, and seems to be very generallyrecognised by geologists. 
It is therefore unnecessary for me to dwell upon it at length. I 
may, however, remark that it is very difficult to make direct 
experiments on the behaviour of mineral substances in this respect ; 
and the theory of such behaviour is apparently not yet sufficiently 
advanced to make it possible to predict what it would be for the 
silicates from any of their other more easily measureable chemical 
or physical properties.’ 
I am at present concerned with a more physical aspect of the 
theory which then ieads me to the discussion of some interesting 
chemical measurements. So long ago as 1850 (the year after 
Professor James Thomson’s’? famous development of the theory 
of the influence of pressure on the melting-point of ice) Bunsen® 
undertook a series of experiments on the influence of pressure on 
the melting-points of spermaceti and paraftin, for the express 
purpose of illustrating the crystallization of igneous rocks. Ihave 
plotted his results, which are very interesting (see fig. 8, p. 402). 
They show that while at ordinary atmospheric pressure spermaceti 
consolidates at a higher temperature than paraflin, yet at pressures 
above ninety-five atmospheres the order of consolidation would be 
reversed. At ninety-five atmospheres the melting-points of the 
two bodies are identical, and so they should consolidate simul- 
taneeously and exhibit a ‘‘ graphic” structure. The great master 
chemist then modestly draws his conclusion :—“Obgleich das 
physikalische Gesetz der Abhangigkeit des Schmelzpunktes vom 
Druck aus diesen wenigen vorldufigen Versuchen nicht einmal 
1 8th Aug., 1901.—NSince the above was written, a Paper on the present subject by 
Prof. W. J. Sollas, r.x.s., has appeared in the Geol. Mag. for July, 1900, in which the 
author dwells upon the possible importance of included water. The frequent secondary 
origin of suchinclusions, as well asthe small and variable quantity of the included water 
in similar rocks, makes it difficult to draw general conclusions as to its influence on 
the order of crystallization in actual rocks. Cf. J. Morozewicz, Tschermak’s Min 
Mittheil. xviii. (1898) 1-90, 105-240, and Min. Mag. xii. p. 313-315. See also A. 
Lagorio, id. viii., p. 421 (1887). 
2 James Thomson, Trans. Roy. Soc., Edinb., vol. xvi. (1849), p. 575. 
3 R. Bunsen, ‘‘ Ueber den Einfluss des Drucks auf die chemische Natur der plu- 
tonischen Gesteine,’”’ Pogg. Ann., 81 (1850), p. 562. 
2F 2 
