Professor T. G. Bonney — Rocks from Kimherley. 501 



4. Conclusion. 



I believe, as T have already stated/ that the diamonds vp-ere not 

 produced where they are now discovered, but, like the conspicuous 

 olivine, pyroxenes, mica, garnets, and iron-oxides, had their origin 

 at a much greater depth in the earth's crust. Thus I think it 

 improbable that the carbon was obtained from the Karoo Shales, 

 In other words, I consider the diamonds, whatever may have been 

 their past history, to have come from the crystalline floor - on 

 vi'hich these shales were deposited, or to be at any rate Pre-Triassic 

 in age. I see no reason why carbon should not be present in the 

 earth's magma, whether it afterwards crystallized in a peridotite 

 or in native iron.^ 



Volcanic action probably began before the end of the Karoo 

 epoch, as the great sheet of so-called melaphyre seems to be 

 intercalated with its higher deposits,* and some of the dykes may be 

 approximately coeval with it. Afterwards (was it because of this 

 impenetrable cover of lava ?) the pipes were formed, probably in 

 a short time, by a series of great explosions, caused by gases 

 accumulated at considerable depth. These sent the outer part of 

 the crystalline floor, including the diamonds, as well as the 

 overlying Karoo shales, flying shattered into the air. Cones and 

 craters may have been built up above these pipes, but I expect that, 

 like some of those in the Eifel, they were never high, and that the 

 shattered material mostly fell back, after a few explosions, into 

 the pipes and filled them like the " necks " in Scotland. Then 

 the volcano, after some more lava had struggled to reach the surface 

 (forming, at any rate, the dykes in the necks), passed into a solfatara 

 stage, during which the numerous secondary changes mentioned 

 above were produced, the carbonates were deposited, and the minor 

 structure of the mass was obscured. Afterwards ordinary meteoric 

 agencies began to work ; water percolating from above still further 

 affected the mass, more especially in its upper parts, producing 

 the " yellow ground " and the " sott blue " and depositing tufa. 



The result of this examination is to confirm me in my opinion that 

 the diamantiferous rock — the so-called Kimberlite — is a volcanic 

 breccia rather than a peculiar form of peridotite. While it is 

 difficult, owing to secondary changes, to demonstrate the fragmental 

 character of the matrix, I have failed to identify it with any form of 

 peridotite (or serpentine) known to me, and I cannot understand 



1 Geol. Mag. 1895, p. 501. 



^ It must be remembered that pyroxenite, troktolite, and gabbro are frequently 

 found associated with peridotites and are generally intrusive in them. 



•* That graphite, cliftonite, carbonado, and diamond occur in the native iron, is, of 

 course, well known : see Fletcher, Min. Mag., vii, 121. Meteorites vary from 

 almost pure native iron to peridotites: see Wadsworth, " Lithological Studies," 

 ch. ii, and Crookes, Nature, Ivi, p. 325. Native iron is well known to occur in 

 some basalts. 



* According to Sir W. Crookes, F.E.S., "The Diamond Mines of Kimberley" 

 (Lecture at Imperial Institute), the section is as follows : (1) soil, ii) basalt, 

 much decomposed, 20-90', (3) black shale (combustible), 200'-25u', (i) "mela- 

 phyre," about 4U0', (5) quartzite, about 400,' (6) variable shales. 



