LACCOLITES AND BYSMALITES. 189 



The only substance which is notably in greater quantities in 

 the dolerite is the iron, and this correspondence of the chemical 

 composition of shales with dolerite. to my mind, explains the 

 predilection of dolerite for intruding itself into shales. But in 

 any body of shales, unless one is dealing with the vast palaeozoic 

 slate formations of Europe, there are always sandy beds, in 

 which the percentage of silica rises to 80 or 90 %, and generally 

 one may state that the average composition of a varied slate and 

 sandstone formation, such as the Beaufort beds, for instance, 

 contains nearer 70 % silica than 50 %. When dolerite, there- 

 fore, invades an area of slate it finds a rock not very different 

 in composition to itself, and a small amount of original liquid 

 material, by incorporating the country rock, may spread out 

 and occupy a very considerable space. Nevertheless, some new 

 material must be brought in, and consequently some of the sub- 

 stance of the absorbed slates must be passed away. From the 

 figures given above, this will be principally silica, and its only 

 channel of escape is the supply dyke. A rising column of liquid 

 igneous rock must, therefore, function as efferent as well as an 

 afferent mass, that is to say, the acid waste products of the 

 absorption of the sediments must diffuse downwards through the 

 up-welling basic magma. 



In the deeper portions of the crust the principal massifs of 

 igneous rock are granite. If we admit the downward passage 

 of acid material, some reservoirs in the deeper portions of the 

 crust ought to be found filled with igneous rock of acid compo- 

 sition ; some of the granite massifs, therefore, should be of such 

 an origin. 



In the south-west of Cape Province the overlying Karroo 

 sediments have been removed by denudation, and also, to a 

 certain extent, the Cape Formation. The latter is p'-edominantlv 

 an arenaceous formation, and hence should not contain much 

 injected igneous rocks ; but the shales below the Table Mountain 

 sandstone are very freely injected with granite, and the massifs 

 are exposed in Table Mountain. Paarl, Malmesbury, Robertson, 

 George and elsewhere. Those which owe their origin to the 

 absorption of strata and reception from above of the residual 

 acid material from the basic intrusions I propose to call bysma- 

 lites. The term bysmalite was proposed by Iddings in 1898 for 

 intrusions of magma which had been forced vertically through 

 other rocks and solidified in the form of plugs. They are laccolites 

 exaggerated in the vertical dimension and. according to Iddings. 

 accompanied by vertical faulting of the containing rocks. The 

 type is taken from Mount Holmes in the Yellowstone Park. The 

 only reason that Iddings adduces for the faulting is that the 

 beds abut abruptly on the igneous rock, and. denying that the 

 igneous rocks can have melted out their cavity, the onlv alterna- 

 tive is to bring in faults to explain the phenomenon. In the 

 basic intrusions, however, we can everywhere see in South Africa 

 that absorption has taken place; therefore, a granite massif with 

 steep sides and the strata abutting against it would be a hysmalite. 



