Alexander Scott — Saturation of Minerals. 321 



The coinpanitive rarit}' of leucite may be explained in a similar 

 ■way. This mineral seems to form only in a small temperature 

 interval,' and if, by undercooling or otherwise, crystallization takes 

 place outside this interval, the final product is a mixture of orthoclase 

 and neplieline. This may also be expressed by saying that the 

 reaction kaliophilite + orthoclase ^=^ leucite 



is revei'sible, and proceeds towards the formation of leucite onlv 

 through a certain temperature range. 



In this case, the effect of undercooling is not merely physical but 

 chemical also, and is therefore analogous to the phenomena observed 

 in the cadmium-antimony alloys.- A mixture of equal atomic 

 percentages of cadmium and antimony, cooling from the fused state, 

 deposits crystals of a compound Cd Sb, if the melt is inoculated; 

 otherwise, without inoculation, undercooling takes place and a mixture 

 of antimony and a compound C'dg Sbg separates. The freezing-point 

 curve of the metastable system therefore differs from that obtained 

 under conditions of stable equilibrium. The system albite-(soda)- 

 nepheline is probably similar, save that the compound, analcite, 

 requires the presence of water before it is formed, in addition to 

 definite conditions of temperature and pressure. 



Undercooling, however, may take place in the solid state as well 

 as in the liquid state, and involve the presence of unstable or 

 metastable phases. Instances of this phenomenon are common amongst 

 the rock-forming minerals, silica itself being a good example. 

 According to recent work,^ /j-cristobalite is the modification of silica 

 stable above 1,470°; /3-tridymite, between 870° and 1,470° ; /i-quartz 

 between 575° and 870°; and a-quartz below 575°. The occurrence 

 of tridymite in certain effusive rocks, such as the trachyte of 

 Drachenfels, is well known, and it will be obvious from the above 

 that the mineral must be an unstable phase. Its origin may be due 

 to crystallization from the magma at a temperature above the 

 transition-point, and then owing to the rapid cooling, the transforma- 

 tion to quartz maj^ be inhibited. It is possible, however, as Feniier* 

 has pointed out, that crystallization may have occurred below 870°, 

 and that an unstable form has developed in accordance with Ostwald's 

 rule, that the most stable phase may not be attained directly, but that 

 phases of intermediate stability may first form. As in the first case, 

 a rapid fall to ordinary temperatures might greatly reduce the velocity 

 of inversion. There are apparent exceptions to Ostwald's rule, but 

 it is possible that these may be explained by the thermodynamic 

 potential of the liquid at the given temperature being lower than, or 

 nearly equal to, the potential of the modifications which fail to appear. 

 In the case of tridymite, quartz, etc., the change is apparently 

 physical and is confined to a polymorphic transformation. 



The hornblendes and pyroxenes may constitute another exam])le 

 ■of the same type. While there is still some obscurity regarding the 



J_ Cf. Washington, Journ. GeoL, xv, pp. 257-79, .357-9-5, 1907. 

 " Kurnakov & Konstantinov, Zeit. anorg. Chem., Iviii, pp. 12-'22, 1908. 

 ' Fenner, " The Stability Eelations of the Silica Minerals " : Amer. Journ. 

 Sci., ser. IV, xxxvi, pp. 331-84, 1913. 

 * Fenner, loc. cit., p. 342. 



DECADE VI. — VOL. I. — NO. VII. 21 



