296 R. H. Rastall — Differentiation and Ore-deposits. 



Thus it is obvious that when a partial magma has reached its final 

 resting-place and is there crystallizing, there must be a more or less 

 complete separation of metallic constituents. This takes place in 

 two stages, first a concentration in the last residue of the crystallizing 

 magma, and then a translation along with this residue into dykes, 

 veins, and lodes. 



This process is best known and the relations are most clear in the 

 case of the acid magmas, but the same thing occurs in the case of 

 basic intrusions, with one important difference. The basic magma 

 is less rich in volatile constituents, fluorine and boron being absent, 

 and for some reason not yet understood water is apparently less 

 abundant. Hence there is less tendency for the metals to segregate 

 into dykes and veins. More commonly they remain more or less 

 evenly disseminated like chromium and platinum, or sink to the 

 bottom like the nickeliferous sulphides. It is worth noting that 

 chromium and platinum are most abundant in the ultrabasic 

 peridotites, which are themselves to be regarded as differentiates 

 of the larger basic magmas. 



We are thus led to recognize two distinct types of ore-deposits 

 of direct magmatic origin, formed by differentiation, namely, the 

 segregations within the mass of the magma itself, such as the 

 nickeliferous sulphides of Sudbury or the iron-ores of Kirunavaara, 

 where there has been only separation by sinking within the basin, 

 and secondly, the fissure deposits, dykes, and veins, formed by trans- 

 lation of metalliferous portions, usually upwards or laterally, and 

 frequently extending beyond the boundaries of the basin ; the 

 clearest examples of this type are afforded by the tin and tungsten 

 deposits, while a similar origin may be postulated for a great number 

 of other metallic ores, including copper, lead, zinc, arsenic, antimony, 

 bismuth, mercury, silver, and gold. It appears probable that as 

 a result of extended study it will become possible to draw up 

 further generalizations, by assigning each metal or group of 

 metals to its own petrographic rock-type. Thus it appears that 

 gold is specially characteristic of the monzonite group, while the rare 

 earths are concentrated into highly alkaline partial magmas. 



With regard to the actual mechanism of differentiation, the subject 

 of so much discussion of late years among petrologists, the matter 

 is of minor importance ; it is the effects with which we are really 

 concerned. All the facts go to show that some sort of splitting of 

 magmas must have occurred ; the question of how or why is mainly 

 of academic interest. The chief battle at present is between the 

 advocates of differentiation by diffusion before formation of solid 

 crystals, and differentiation by gravity after the crystals are formed, 

 while the immiscible liquid theory comes in a bad third. Now it is 

 obvious that partial magmas which are still liquid after the 

 differentiation cannot have been formed by sinking of solid crystals, 

 or, at any rate, both fractions cannot. The upper fraction might, 

 of course, beconae impoverished in certain constitueuts by sinking 



