F. P. Mennell — Archcean Sirafigrcqjh//. 261 



•evidence adduced by me to the contrary.^ The dome-like structure 

 of the schists round most of the granite masses is readily accounted 

 for on the idea of the greater age of the latter. What has, however, 

 given rise to the greatest confusion is the fact that at their roots 

 granite masses do not show those features which we have learnt to 

 regard as characteristic of intrusion where we have seen them, as is 

 generally the case in Europe, invading normal sediments. Actual 

 dykes are rare, and do not extend far from the main mass of the 

 rock. The characteristic feature of the invasion of crystalline 

 schists is in fact the production by ' lit-par-lit injection ' of those 

 very rocks which have aroused the greatest amount of controversy. 

 Once this fact is realised, much becomes clear that before was very 

 far from it. The scantiness of dyke-like protrusions is readily 

 explained. At great depths the formation of open fissures is 

 scarcely possible ; eruptive activity is naturally prominent only in 

 the upper part of a mass. The offshoots from the deep-seated 

 portions must corrode a way into the surrounding rocks, or insinuate 

 their material between their laminae. The last process is the usual 

 one, and even the few large dykes that are seen are mere feeders 

 by which the injection is carried on, and which soon become 

 exhausted in the process. They therefore scarcely ever extend 

 beyond the zone of injection 'lit par lit,' though as this may be 

 over a mile across, that would not prevent them being prominent, 

 and their extreme rarity is rendered the more evident. 



The section at Figtree referred to above is one of those rare 

 occurrences which outdo textbook diagrams in clearness. There 

 are excellent exposures along the spruit-beds, and we may see large 

 dykes running into the schists for hundreds of yards, as well as 

 every intermediate stage between such features and the insinuation 

 on a microscopic scale of granitic material between the foliation 

 planes of the schists. Sometimes the granitic and sometimes the 

 schistose material has, owing to slight movements, been broken up 

 into discontinuous threads. Thus we see one rock forming a matrix 

 enclosing a series of lenticles of the other, and subsequent diffusion 

 of material, or possibly actual melting, has in some cases also led to 

 the formation of typical banded gneisses, while epidosites are also 

 found, due no doubt to what Dr. Callaway calls ' secondary 

 injection.' When once a section like that described has been seen, 

 it is easy to interpret more obscure ones in other places. Not that 

 other sections are necessarily obscure at all; in fact, when the 

 unfamiliarity has been overcome, it is an easy matter from the 

 nature of the rocks to predict the proximity of granite when 

 crossing a metamorphic area, or of schists when traversing an 

 igneous one. In a number of localities it is very interesting to 

 trace the gradual changes that take place. Starting, for instance, 

 from beyond the Eifle Kopje at Bulawayo, we begin with coarsely 

 crystalline, unfoliated ' epidiorites,' which, in spite of amphi- 

 bolisation, etc., retain an ophitic structure, which puts their original 

 nature beyond doubt. Approaching the granite going nearly west, 



1 " Geology of South Africa," 1905, p. 98. 



