78 E. H. L. SCHWARZ 
described Lystrosaurus, which occurs on a higher horizon, but still 
within the Beaufort Series proper, so that these beds must occur in a 
strip running more or less north from East London, where again 
reptiles of the same horizon occur; but their development is not 
continuous. In Umsikaba (Flagstaff), in Eastern Pondoland, Dr. 
Rogers and myself found the Umsikaba Beds (the local facies of the 
Ecca Beds) directly followed by the Idutywa Beds. At the time, 
not knowing that the Kentani Beds wedged out, we placed the 
Idutywa Beds below the Dicynodon or Kentani Beds, thinking that 
they represented the Pareiasaurus—zone of the Beaufort Series; 
this was a natural mistake when working upward in the series, as 
there is a complete conformity between the Dwyka and Umsikaba 
(Ecca) Series and the Umsikaba and Idutywa Series. Later, how- 
ever, working downward, we found the Molteno Beds of the 
Stormberg System lying directly on the Idutywa Beds. 
Western Karroo Eastern Districts Pondoland Transvaal 
Idutywa Series Idutywa Series | Idutywa Series High Veld Series (Coal 
Measure Series) 
Beaufort Series | Kentani Series | Wanting or only local- | Wanting 
ly developed 
Ecca Series Ecca Series Umsikaba Series Wanting 
Dwyka Series Dwyka Series Dwyka Series Dwyka Series. — 
PER SALTEM CONFORMITY 
Such a conformity as that represented by the Umsikaba Series 
being followed by the Idutywa Series I will call a per saltem con- 
formity. It is due to want of supply of sediment during some 
lengthened period. The most conspicuous example of it that we 
have in South Africa is the conformity between the Bokkeveld 
Series (lowermost Devonian) and the Witteberg Series (Lower 
Carboniferous), the whole of the Upper and Middle and most of the 
Lower Devonian being unrepresented. On the supposition that 
the fossils in South Africa have the same value as time markers as 
those in the northern hemisphere—which is a debatable point, but 
one which seems to become more definite as the material accumu- 
lates and the study of it advances—this conformity between the 
Witteberg and Bokkeveld Series is explained by the shore retreating 
