GABBROS ETC. IN SCOTLAND AND IRELAND. 79 



dation has taken place have certainly had not a little to do with the 

 order of crystallization of the minerals of the rock. 



Before any satisfactory conclusion can be arrived at concerning the 

 order in which the several minerals of a rock have crystallized from 

 the magma, we must allow for two possible sources of error in 

 reasonings of this kind: — 



First we must determine with accuracy which are the original 

 minerals of the rock, and carefully distinguish them from such as 

 are of secondary origin. 



Secondly it must be borne in mind, as has been so justly insisted 

 upon by MM. Michel-Levy and Fouque, that there may be several 

 distinct periods of consolidation of the minerals in a rock, and that 

 some crystals may even be derived, ready formed, from a foreign 

 source. 



Bearing these considerations in mind, I have been able to arrive 

 at the following conclusions with regard to the rocks we are now 

 studying : — 



The oxides of the iron-group of metals {iron^ manganese, chromium^ 

 nickel, and cobalt) behave quite differently in basic magmas which 

 have consolidated slowly and under pressure, and in the same magmas 

 cooling at the surface. This is shown by the fact that in the most 

 deeply seated magmas there are no original spinellids (magnetite, 

 chromite, or picotite), but the whole of the oxides of the iron-group 

 exist in union with silica, either as pyroxenes or olivine. The careful 

 study of the perfectly crystalline types of all these rocks from 

 different portions of the area under consideration shows that mag- 

 netite, when present in them, is always a secondary product resulting 

 from the " schillerization " of the olivine or the pyroxenes. I find 

 that Prof, von Lasaulx, from his examination of a series of the 

 Carlingford gabbros in 1878, was convinced that the magnetite in 

 his specimens was of secondary origin *. 



And a critical examination of the whole series of these rocks 

 shows that as the cooling becomes slower and the pressure less, the 

 quantity of the oxides of iron separating as magnetite progressively 

 increases. In some of the less highly crystalline gabbros and in the 

 dolerites, small quantities of original magnetite can be detected ; in 

 the basalt it increases in amount ; in the magma-basalt the whole 

 base of the rock is thickly sown with skeleton-crystals or dust of 

 magnetite ; and in the tachjiytes it sometimes becomes so abundant 

 as to render the rock completely opaque even in the thinnest 

 sections. 



It follows, as a necessary coroUary from thi§ proposition, that the 

 pyroxenes and olivines which separate from a magma when it is 

 deep-seated must be more highly ferriferous than the same species 

 of minerals crystallizing from the same magma nearer the surface. 

 And there are many facts which appear to support this conclusion. 

 It is the highly ferriferous character of the deep-seated olivines 

 which seems to facilitate their alteration into opaque masses by schil- 

 lerization. Another conclusion which may be drawn from the study 

 * Tschermak's Min. und Petr. Mitth. vol. i. (1878), p. 429. 



