Great Dyke of Norite of Southern Bhoclesia. 
5 
■of the intrusion. He further states (p. 51) that " The exact nature of the 
Great Dyke is by no means clear. In the bed of tlie Umgezi river, which 
is fairly deeply incised, pink granite is exposed for a distance of fully 
a quarter of a mile beyond the line of contact between granite and 
pyroxenite on the western portion of the farm Umgezi. This circum- 
stance, taken in conjunction with the fact that in the western half of the 
intrusion at any rate the dip of the layers of pyroxenite and peridotite 
appears everywhere to be to the south-south-east, suggests that the Dyke 
is in reality a huge sill dipping at a low angle to the south-south-east. 
This view, however, receives no support whatsoever from the relationship 
between pyroxenite and granite along the eastern margin of the intrusion 
or from the relationship between pyroxenite and norite on the Springs and 
Sandeman's farm, for the contact between the last-mentioned rocks 
appears to be vertical, and to the writer it appears more probable that 
in the area under review the Great Dyke is of the nature of a very 
elongated laccolite. Further investigation, however, is urgently needed 
on this point." 
It is advisable therefore to state fully the evidence obtained at 
Selukwe with a view to substantiating or otherwise the opinion of the 
present writer. 
It may be pointed out that in general the intrusion has many of the 
characters of the dyke ; thus its rectilinearity of outcrop, the parallelism of 
its sides and fairly constant width (which is small in comparison with its 
longitudinal extension), together with the symmetrical disposition of the 
variations, indicate a dyke rather than a sill or laccolite. 
It must be admitted, however, that the Selukwe portion is poorly 
exposed. The junctions of one rock with another are rarely seen, and 
the actual edge of the intrusion was only visible in five or six exposures, 
and these with one exception (Gwania) are on the eastern side. 
Each of these exposures of the edge of the intrusion is steeply inclined 
or vertical, and there was no modification of the rock noted that is indica- 
tive of slow cooling ; this fact is often interpreted in favour rather of the 
dyke habit than of the inclined sheet or laccolite. Several of these edge 
exposures exhibit indications of a rude horizontal pillar structure. 
No V-ing of the edges was detected either up or down in the stream 
sections either on the east or the west (but in this connection it must be 
stated that the method of simple plane-tabling in an area where exposures 
are infrequent and the topography inaccurately represented on the map 
used is not to be expected to afford the indications named). Further, 
although the norite appears to be confined to the high ground and the 
pyroxenite to the low ground, in both instances the opposite obtains and 
indicates vertical junctions of the two rocks. Thus in several places 
the Tebekwe rivers expose on the one bank a continuous clilf of norite 
