112 REPORTS ON THE STATE OF SOIENCE.—1917. 
southern Ecca beds were being laid down. At present the north- 
eastern limit of the Upper Dwyka shales is not known, but some beds 
characteristic of those shales in the Cape and 8.W. Protectorate are 
missing in the areas yet mapped in the Transvaal and Natal. 
It is likely that the term ‘ Dwyka’ will be found suitable for the 
‘glacial conglomerate’ in the Transvaal also, and it is already used 
there by several geologists. 
There are different opinions held as to whether the Dwyka series 
is of Upper Carboniferous or of Lower Permian age, but the few 
fossils found in it are not in conflict with the former view. 
The Ecca Series. 
The Ecca series is best known in the Cape and Natal, and the in- 
formation as to its distribution in the Orange Free State and Transvaal 
is not sufficient to decide disputed points such as whether the Ver- 
eeniging coal-beds belong to the Ecca or Beaufort. 
The fossils from undoubted Hcca beds, i.e., beds above the Upper 
Dwyka shales and below those in which there are many genera of Dinoce- 
phalians and Therocephalians, are but few in number. The thickness 
of the series at its maximum is 6,000 feet, and the fossils are 
chiefly fragments of Gangamopteris, Glossopteris, Phyllotheca and 
wood, all of which, except perhaps the first named, occur in the Beaufort 
beds also. The reptiles are very little known, Archazosuchus and Ecca- 
saurus are fragmentary specimens. Whether the so-called Ecca beds 
of Worcester with Gangamopteris really belong to this group is un- 
certain, for superficial deposits conceal the passage down into the 
Dwyka, outcrops of which lie four miles away. 
It thus happens that in the absence of reptilian-bearing Beaufort 
beds the determination of the Ecca is practically impossible in an 
area which has not been connected by mapping with a better-known 
district. 
The Beaufort Beds. 
The Beaufort beds contain many reptilian fossils in certain areas, 
but there appears to be a lack of them elsewhere, and the few plants 
found in the lower portion of the series are not known to be characteris- 
tic. The Beaufort beds have been divided into three sub-groups which 
are again sub-divided :— 
Cynognathus zone. 
Upper or Burghersdorp beds. Procolophon zone. 
Middle Beaufort beds. Lystrosaurus zone. 
Cistecephalus zone. 
Lower Beaufort beds. Endothiodon zone. 
Tapinocephaloid zone.* 
Owing to the comparative rarity of the fossil reptiles and the diffi- 
culty of determinations of individual bones other than parts of the 
skull, these fossils are not well suited for the purpose of the field 
4 This term is substituted for Pareiasaurus zone on account of the revisicn 
of the genus Pareiasaurus by Mr. Watson, according to which no species of it are 
left in the beds concerned. ? 
