VOLCANIC PIPES 361 



an ultrabasic rock and appears to be very closely allied 

 both to the limburgites and to the melilite-basalts. 



Much of the Kimberley blue-ground, indeed, contains 

 porphyritic crystals of olivine, so that Carvill Lewis 

 was led to regard it as a true igneous lava and not as a 

 mud or ash, but it seems more justifiable to look upon 

 the contents of the pipes as breccias derived from the 

 explosive disintegration of bodies of ultrabasic lava ; 

 another effect of the explosions was to break up more 

 or less completely certain basic and ultrabasic holo- 

 cry stalline rocks and to throw the rock and mineral 

 fragments thus obtained up the channels opened by the 

 explosions mingled with the lava in a plastic or solid 

 state. In the upper portions of the pipes this eruptive 

 material received further additions from the walls, and 

 there appears to have been a great amount of stirring 

 up and mixing of the heterogeneous mass. Not only 

 do we find material obviously derived from great depths, 

 but, on the other hand, in the deepest workings of the 

 Kimberley Mine, more than 2,000 feet below the base of 

 the Karroo formation, fragments of Karroo shale form 

 an abundant constituent of the breccia. 



The occurrence of the diamond as a constituent of 

 some of the breccias has been the cause of far wider 

 interest in the pipes than would otherwise have been 

 the case. For many years the diamond was thought 

 to have originated by the crystallisation of carbon 

 derived from the carbonaceous shales surrounding the 

 pipes, the mines of the Kimberley group being at the 

 time the only examples known. Ultimately this hy- 

 pothesis proved untenable^ and the discovery of well- 



