64 PROF. J. W. JTJDD ON" TEETIAET 



anorthite and labradorite *. When there have been two distinct 

 periods of consolidation for the felspar-crystals, the oldest are 

 always the most basic in composition and approach anorthite. 



The augites seldom, if ever, exhibit evidence of having separated 

 at two different periods. But in different varieties of the rock they 

 present a wide range in chemical composition, from diopsides and 

 chrome-diopsides to highly ferriferous augites, as is shown by the 

 series of analyses given t. 



The olivines exhibit the same wide range of variations in com- 

 position, from the iron olivines (fayalite) to the lime-iron olivines. 



Seeing that the oxides of iron do not separate as magnetite in 

 the most highly crystalline forms of these rocks, it follows that the 

 olivines and pyroxenes of the gabbros should be more highly ferrife- 

 rous than those of the dolerites and basalts. Observation fully con- 

 firms this inference, the most highly ferriferous varieties of augite, 

 enstatite, and olivine being found in the more perfectly crystalline 

 varieties (gabbros). 



Some of the accessory minerals, such as apatite and garnet, must 

 be regarded as additional constituents ; others clearly replace to a 

 certain extent one or other of the essential constituents. 



The most striking case of this substitution of one mineral for 

 another in these rocks, is the appearance of enstatite in the place of 

 augite. The enstatite in every case plays exactly the same role in the 

 rock as the augite. More than this, a highly ferriferous augite is 

 usually replaced by a highly ferriferous enstatite, and a less ferri- 

 ferous augite by a less ferriferous enstatite. Indeed the relations 

 between those two minerals in the rocks we are considering seems 



* To the series of analyses quoted iu my former paper, I am able to add 

 the following : — Mr. J. F. Brooks, at my request, isolated a sample of the felspar 

 from a -very typical specimen of the gabbro of Beinn More in Mull. This 

 work was executed in the Greological Laborator ^ of the Normal School of Science 

 and ifeoyal School of Mines, by the removal of the heavy ferro-magnesian 

 silicates from the powdered rock with the electro-magnet and by the use of 

 Klein's borotungstate solution. A very satisfactory specimen of the felspar was 

 obtained in this way, which, on analysis, gave the following result : — 



Silica 47-90 



Alumina 31"30 



Iron oxides Trace 



Magnesia 1'16 



Lime 11-22 



Soda 3-96 



Potash -98 



Loss on ignition 1-54 



Mr. Brooks informs me that the deficiency in this analysis probably arises 

 from a slight loss in the silica and magnesia ; the titanic acid was not separately 

 estimated. It will be seen that the felspar is a labradorite. The percentage 

 of magnesia is accounted for by the quantity of serpentinous matter which has 

 penetrated into the cracks of the crystals, as is seen to be the case when thin 

 sections are examined under the microscope. 



t Quart. Journ. Geol. Soc. vol. xH. (1885), p. 368. 



