336 DE. F. DIXET OX THE [vol. Ixxviii, 



is probably due to the imniiscil)ility of these compoiuicis with the 

 common akmiinous silicates.^ 



The actual order in which the iiiinerals ciystallized when not 

 too strongh' affected by eutectic proportions was as follows : — 

 (1) apatite and zircon ; (2) magnetite, olivine, and pyroxenes in 

 small amount; (3) labradorite ; (4) most of the olivine ; (5) most 

 of the pyroxenes; (6) most of the magnetite; (7) biotite ; 

 (8) micropegmatite and quartz. This order agrees closely Avith 

 that of the Duluth gabbro.- The periods of crystallization, other 

 than (1) and (8), were not always sharply delined. however, and a 

 considerable amount of overlapping ensued : for instance, the 

 magnetite, although occurring principally as in the above list, 

 nevertheless kept up a certain amount of crystallization all through 

 th.e periods, almost to the end. The magnetite of certain basalts 

 in Franz Josef Land behaves in a similar way, as was pointed out 

 by Sir Jethro Teall as long ago as 1897 (Geol. Mag. p. 554,). 



A little biotite formed on some of the earlier crvstals of mag- 

 netite, and, in general, diallage ])receded hvpersthene. 



Finally, it is worthy of note that, although crystallization in the 

 principal member of the norite-complex did not follow the common 

 sequence, the different members of the complex were nevertheless 

 intiTided in the normal order : that is, that of decreasing basicity. 



(6) Contact-Metamorphism within the Complex. 



Some interesting contact-effects have been observed as a result 

 of the invasion of one member of the complex by another. These 

 effects scenerally took the form of corrosion and recrvstalhzation. 

 Sometimes, however, small quantities of hornblende and biotite 

 were developed ; but, apart from these, no new minerals were 

 produced. This is in the main what would be expected, since the 

 composition of any one principal intrusion did not differ very 

 much from that of another. 



Such changes as have taken place at the Sierra -Leone contacts 

 are more closely paralleled in certain cases described by Prof. 

 Lacroix ^ from the Central Plateau of France and elsewhere, 

 rather than in any of the British occurrences. Most of the latter 

 liefer to the alteration by plutonic masses, not of basic plutonic 

 rocks, but only of basic lavas; and it is imcertain in most instances 

 ihow far secondary products had developed in the lavas before they 

 were invaded by the new magmas.^ Consequently, these cases 

 'Cannot often be taken as good examples of the thermal metamor- 

 ■phism of basic rocks. The British occurrences to which I would 



^ A. Harker, ' The Xatiiral History of Igneous Eocks' 1909, p. 200 ; and 

 .J. Y. Elsden, ' Principles of Chemical Geology ' 1910. p. 135. 



•2 M. L. XebeL Econ. Geol. vol. xiv (1919)'p. 372. 



^ ' Les Enclaves des Eoches Yolcaniones ' Ann. Acad. Macon, ser. 2. vol. x 

 (1893). 



^ See A. Harker & J. E. Marr. • Supplementary Xotes on the Metamorphic 

 JRocks .around the Shap Granite' Q. J. G. S. vol.'xlix (1893) p. 360. 



