part 4] i^OEiTE OF sieeea leone. 335 



the three minerals (felspar, magnetite, and pyroxene) will proceed 

 to ciystallize together in eutectic proportions and at a stationary 

 temperature. 



Similarlj^ it will be seen that in the cooling of a magma of 

 composition X, magnetite would crystallize out until the residue 

 attained the composition Y, wdien felspar would crystallize with 

 the magnetite ; the remaining magma would finally attain the 

 composition E, at Avhich the ternary eutectic (magnetite, felspar, 

 augite) would be formed, as before. 



The ternary eutectic of a pj^'oxene C^ an olivine D, and 

 magnetite B, or of any other of the ternary sj^stems described 

 above, could also be expressed by this method. 



(5) The Order of CrystallizatioD. 



It is obvious from the fores^oino- remarks that the order of 

 crystallization in the norite-magma was far from being in 

 accordance with the law, named after E-osenbusch, which commonly 

 holds in the case of basic magmas. The principal departures from 

 this law rest in the early crystallization of the labradorite, the 

 prolonged and generally late crystallization of the iron-oi'es, 

 the variable crystallization of the olivine and pyroxenes, and the 

 late crystallization of the biotite. The order of crj'-stallization 

 followed the law, however, in the early precipitation of the apatite 

 and zircon. Prof. J. H. L. Yogt's more simple law, demonstrated 

 by his researches on slag,i that the order of crystallization is 

 determined by the relative proportions of the several minerals 

 present as compared with the eutectic proportions, seems to have 

 been the leading principle. This is confirmed by the occasional 

 development of a quaternary intergrowth which closely approaches 

 the theoretical eutectic of the magma. With regard to similar 

 conditions in the vSkiddaw Granite (see above, p. 331) Dr. R. H. 

 E-astall concludes : — 



' It appears to follow that the order of crystallization was that which finally 

 led to a eutectic cojuposition which expressed itself as a graphic intergrowth 

 of a varying number of components. The large and comparatively pure 

 phenocrysts represent the excess of certain components over this eutectic 

 ratio.' (Q. J. G. S. vol. Ixvi, 1910, p. 121.) 



Nevertheless, certain anomalies in the crystallization of the 

 norite were possibly due to the influence of superfusion and 

 viscosity ; superfusion itself may cause not only reversals in 

 the order of crystallization, but also may result in a marked 

 sequence of separation in place of the simultaneous crystallization 

 of the eutectic. 2 The early precipitation of the apatite and zircon 



1 ' Die Silikatschmelzlosungen ' pt. i (1903) Vidensk. Selsk. Skrifter, 

 Christiania. 



2 W. Meyerhoft'er, ' Schmalzpunkte & Ausscheidungsfolge von Mineralien ' 

 Zeitsohr. Krystallogr. vol. xxxvi (1902) pp. 593 et seqq. ; see also J. V. Elsden, 

 'Principles of Chemical Geology' 1910, pj). 151 & 154. 



