TRANSACTIONS OF SECTION C. 721 



or less baked aiul altered. The occurrence of miuute tourmalines is evidence of 

 fumarole action. 



Tlie microscopical examination supports the geological data in testifying to the 

 igneous and eruptive character of the peridotite, which lies in the neck or vent of 

 an old volcano. 



While belonging to the family of peridotites, this rock is quite distinct in struc- 

 ture and composition from any member of that group heretofore named. It is 

 more basic than the picrite porphyrites, and is not holocrystalline like dunite or 

 saxonite. It is clearly a new rock-type, worthy of a distinctive namo. The name 

 Kimberlite, from the famous locality where "it was first observed, is therefore 

 proposed. 



Kimberlite probably occurs in several places in Europe, certain garnetiferous 

 serpentines belonging here. It is already known at two places in the United 

 States : at Elliott County, Kentucky, and at Syracuse, New Yorlf ; at both of 

 which places it is eruptive and post-carboniferous, similar in structure and com- 

 position to the Kimberley rock. 



At the diamond localities in other parts of the world diamonds are found either 

 in diluvial gravels or in conglomerates of secondary origin, and the original matrix 

 is difficult to discover. Thus, in India and Brazd the diamonds lie in a conglome- 

 rate with other pebbles, and their matrix has not been discovered. Eecent obser- 

 vations in Brazil have proved that it is a mistake to suppose that diamonds occur 

 in itacolumite, specimens supposed to show this association being artificially manu- 

 factured. But at other diamond localities, where the geology of the region is 

 better known than in India or Brazil, the matrix of the diamond may be inferred 

 with some degree of certniuty. Thus, in Borneo, diamonds and platinum occur 

 only in those rivers which drain a serpentine district, and on Tanah Laut they also 

 lie m serpentine. In New South Wales, near each locality where diamonds occur, 

 serpentine also occurs, and is sometimes in contact with carboniferous shales. 

 Platinum, also derived from eruptive serpentine, occurs here with the diamonds. 

 In the Urals, diamonds have been reported from four widely separated locaUties, 

 and at each of these, as shown on Murchison's map, serpentine occurs. At one of 

 the localities the serpentine has been shown to be an altered peridotite. A diamond 

 has been found in Bohemia in a sand containing pyropes, and these pyropes are 

 now known to have been derived from a serpentine altered from a peridotite. In 

 North Carolina a number of diamonds and some platinum have been found in river 

 sands, and that State is distinguished from all others in eastern America by its 

 great beds of peridotite and its abundant serpentine. Finally, in northern 

 California, where diamonds occur plentifully and are associated with platinum, 

 there are great outbursts of post-carboniferous eruptive serpentine, the serpentine 

 being more abundant than elsewhere in North America. At all the localities men- 

 tioned chromic and titanic iron ore occur in the diamond-bearing sand, and both of 

 these minerals are characteristic constituents of serpentine. 



All the fucts thus far collected indicate serpentine, in the form of a decomposed 

 eruptive peridotite, as the original matrix of the diamond. 



4. Ohservations on the EonuUng of Pehhies by Alpine Rivers, ivith a Note on 

 llielr Bmring upon the Origin of the Banter Conglomerate.^ By Pro- 

 fessor T. G. BoNNEY, D.Sc, LL.D., F.B.S., F.G.S. 



The author describes the result of his observations of the rounding of pebbles in 

 various torrents and rivers in the Tyrol and Dauphine, and of the gravels of the 

 Piedmontese and Lombard plains, these lead to the following conclusions, among 

 others: (a) that pebbles are rounded with comparative rapidity when the descent 

 of the stream is rapid, and they are dashed down rockv slopes by a roaring torrent, 

 capable of sleeping along blocks of much greater volume; (b) that pebbles are 

 rounded with comparative slowness when the descent is gentle and the 

 average pace of the river is about adequate to push them along its bed. The rocka 



' Geol. Mag. 

 1887, ^ Q , 



