The Occurrence of Dinosaurs in Bushinanland. 267 
water in the well, so that an examination of the further work below 
ground could not be made. More fragments of bone were found lying 
on the dump, and from these and the bones got in 1913 femurs with 
dinosaurian features have been put together. Mr. Haughton has written 
an account of the bones. 
It being very probable that we have not got to do with an outlier of 
a formation containing dinosaurs covered discordantly by very much 
younger deposits, there remains to be considered whether the bones 
were derived from such a formation which used to exist, or still exists, in 
some part of Bushmanland, and are merely boulders in the bed of the 
valley like the fragments of gneiss and quartz. The bones are neither 
water- worn nor weathered ; their condition differs very greatly from, for 
instance, that of the bones picked up in the Karroo after having been 
released from the Beaufort beds by weathering. The larger bones at 
Kangnas have in some cases been broken in situ and the fragments 
re-cemented by carbonate of lime, but the surfaces are in good condition 
and the ends are not battered. It seemed to me that the bones had been 
buried in the sandy mud, and that the history of the valley since then has 
been a more or less uniform process of in-filling by the material washed 
ajnd blown into it. 
So far as is known, dinosaurs became extinct in Cretaceous times, 
though in view of the discussions on the age of the later dinosaur-bearing 
beds in America that question is still an open one ; so the age of the bones 
is not known, indeed the history of dinosaurs in Africa is at present most 
fragmentary. However, these bones must be very old in a geological 
sense, and their occurrence at Kangnas leads one to suspect that many 
similar finds may be made in the buried valleys of Bushmanland, the 
southern Kalahari, and part of Griqualand West. 
No shells or other remains of animals than the bones have been found 
in the well, and the fossil wood has not yet been examined. 
The nature of the material in the well reminds one very strongly of 
the deposits exposed in the ravines on the eastern flank of Kamies Berg 
and the adjoining part of Bushmanland, where rubble and sandy clay lie 
on the old gneiss, and where there are occasional masses cemented by 
limestone or silica. The occurrence of silicified wood recalls the curious 
deposits at Banker,''' which, however, differ from the Kangnas rock in 
containing ilmenite and opal, and also in the curious bleaching of the 
gneiss and the distinct bedding of the sandy deposit. 
It is important to note that no remains of the Karroo beds occur above 
the gneiss in the Kangnas well, nor were boulders derived from the 
Dwyka tillite found there ; the rock fragments are of local origin. The 
lower part of the Karroo beds probably covered the whole of Bushman- 
* Annual Keport Geological Commission for 1911. 
