114 REPORTS ON THE STATE OF SCIENCE.—1917. 
beds of South Africa are contemporaneous (in a geological sense) with 
those of Bacchus Marsh. The fact that there is a glacial horizon in 
the Upper Marine series in New South Wales some distance above the 
main Lochinvar glacials shows that it is not possible to use the occur- 
rence of boulder beds for a very close correlation. From personal 
observation I fancy that there is a more considerable faunal difference 
between the Lower and Upper Marine series in the Hunter Valley than 
is usually recognised. 
(B.) Questions 1 and 2.—Mr. Etheridge’s first use of the term 
Permo-Carboniferous was for the Gympie of Queensland (=Lower 
Marine) with an admixture of Star (Carboniferous) forms. According 
to Prof. T. W. Edgeworth David, Mr. Etheridge now extends this term 
to include all Australian formations from the bottom of the glacial series 
to the uppermost beds which contain Glossopteris. 
The term Permo-Carboniferous has, however, been incorrectly used 
by many authors to imply a restricted series of beds of either lowest 
Permian or Upper Carboniferous age, such as the reptile-bearing 
Wichita and Clear Fork of Texas, the Artinsk stuffe of Russia, and 
certain parts of the Salt Range succession. In most of these cases it 
has been made use of to escape the difficulty of deciding on a definite 
Permo-Carboniferous boundary in a consecutive series of rocks, a diffi- 
culty which is really exactly doubled by such action. 
In South Africa the series of rocks to which the term could be 
applied are all of land origin, partly lacustrine, partly zolian, and partly 
river deposits, and hence very considerable discontinuities of deposition 
may occur in an apparently conformable series. Bearing this caution 
in mind, the following represents the apparent conditions : 
The Witteberg beds are directly and apparently conformably suc- 
ceeded along the whole southern margin of the Karroo by the Dwyka 
shales, in which are included the glacial beds. Further north the 
Dwyka conglomerates rest on older beds, partly, no doubt, as a result 
of overlap of land deposits from the area of deposition of the Witteberg, 
but apparently also partly owing to real unconformity. The Upper 
Dwyka shales pass into the Ecca, which is directly continued by the 
Beaufort series, which is again conformable to the overlying Stormberg 
beds. 
This statement is founded on the conditions in the area south of 
the Orange River. In the Orange Free State there seem to be very 
large gaps in the series, the Stormberg series resting on the Beaufort 
series some distance below its top, and this in turn directly on the 
Dwyka, or at most separated from it by a very thin Ecca. 
The Beaufort series is divided into the following zones: 
Stormberg series. . Upper Trias—Rhetic. 
Cynognathus beds . : . Middle (? Lower in part) Trias. 
Procolophon beds . : . No direct evidence of age. 
Lystrosaurus beds . J . is % aA 
Cisticephalus beds . : - Upper Zechstein (=Dwina beds). 
Endothicdon beds . ; . No direct evidence of age. 
Tapinocephalus beds. . . ? Lower Zechstein. Evidence from 
(olim Pareiasaurus zone) comparison with copper-bearing 
beds of Orenburg. 
