﻿Vol. 66. J GEOLOGICAL STRUCTURE OF SOUTHERN RHODESIA. 359 



In common with all the other members of the metamorphic group, 

 the great granite-masses have an intrusive relation to the conglo- 

 merates, while the epidiorites may also occur as obvious dykes in 

 the series, as, for example, in the bed of the Hunyani River and at 

 the Eldorado Mine, in both of which cases the dykes are nearly 

 vertical and cut the beds almost at right angles. In no case, 

 moreover, have fragments of the ordinary epidiorites been found 

 among the pebbles at any of the numerous localities that I have 

 visited. Banded Ironstone pebbles, on the other hand, are nearly 

 always to be found, and often predominate over any others, as at 

 the Riverlea Mine. Discordances in the direction of strike, as well 

 as overlaps, point likewise to an unconformity between the Banded 

 Ironstones and the Conglomerate Series. A very interesting point 

 was revealed by the microscopic study of a series of pebbles col- 

 lected at Lomagundi. As already noted, although the present-day 

 exposures of granite are later than these beds, yet granite pebbles 

 are extremely numerous at many localities. They differ from the 

 intrusive granites in several features, notably in never showing 

 microcline, which is usually the dominant felspar and is only 

 locally rare or absent from the intrusive granites. ' Elvan '-like 

 pebbles are also abundant : they indicate that a contact-zone was 

 being denuded, certain actinolite-schists, etc., no doubt representing 

 some of the contact-altered rocks. The Banded Ironstone pebbles 

 are interesting, as showing that the rock had evidently assumed its 

 present features when the Conglomerates were being laid down. 

 Further than this, a specimen after being sliced showed one of the 

 characteristic types of contact-alteration often seen round our 

 normal granites, being changed to a coarsely granular aggregate of 

 magnetite, quartz, and secondary hornblende. It is to be presumed, 

 therefore, that even this early granite which furnished the pebbles 

 was intrusive in the Banded Ironstone Series. It is possible, as I 

 have already suggested, 1 that the gneissic rocks of the Surprise 

 Mine in the Selukwe district may represent a granite of this age, 

 but, if so, it has largely lost its original characters. All supposed 

 records of granites earlier than the schists have been based on 

 errors of observation, just as has become increasingly evident every 

 year in Canada. 



The volcanic rocks associated with the Lomagundi con- 

 glomerates have been mentioned incidentally above, but have not 

 been examined microscopically. They undoubtedly often represent 

 ancient lavas, and occur nearly at every point where the Conglome- 

 rates are developed. They are often highly amygdaloidal, the 

 cavities being filled with calcite, quartz (representing, probably, 

 original chalcedony or agate), or even with felspar, no doubt 

 replacing various zeolites. I have a collection of a variety of types 

 made in 1907 at Belingwe, between the Dobi River and the Native 



1 ' G-eology of Southern Khodesia ' 1904 (Spec. Eep. No. 2, Ehod. Mus.) p. 11. 



