Flora, after which another flora of a different character took its place. 
There is, however, this difference, that in Australia the change appears to 
have taken place at an earlier date, because there we have only a Lower 
Carboniferous flora, after which Glossopteris appears ; while in Africa we 
find still an Upper Carboniferous flora before the deposition of the boulder 
bed came on. 
8. A second bed of conglomerate, which also is said to be due to ice 
action, is mentioned from the so-called Kimberley Shales above the Upper 
Ekka beds. 
9. The Karoo Formation, as it is at present defined, comprises, there- 
fore, the Kimberley Shales, together with the overlying sandstones, representing 
apparently the Koonap, Beaufort, and one portion of the previous Stormhcrg 
beds, where occurred the fossils mentioned by Tate, such as Glossopteris, 
Hubidgea, Vkyllotheca, as well as the animal remains. 
If we now look for analogies we shall find the following : — 
(«) The l)wyka Conglomerate of the Ekka beds has to be correlated 
with the conglomerates and boulder beds at the base of the Talchirs in the 
Salt liauge (below the Froductus limestone), at the base of the Bacchus 
Marsh beds, and below the Newcastle beds, New South Wales. The Dwyka 
Conglomerate is, as we have seen, not older than “Upper Carboniferous” 
lying above strata with coal plants of the type of the Upper Coal Measures 
in Europe. In Australia the deposition seems to have set in somewhat 
earlier, and also in the Salt Mange it appears of somewhat older date, under- 
lying the Froductus limestone, which, as it appears, represents also partly the 
Upper Carboniferous Epoch.* 
(b) What lies directly above the Dwyka Conglomerate, viz., the Upper 
Ekka Shales, must be correlated with the respective beds above the boulder 
beds elsewhere, viz., with the Talchir-Karharbari beds in India, Bacchus 
Marsh beds in Victoria, and Newcastle beds in New South Wales; and the 
age of these would be about Fermian. Equivalents of these would also he 
the upper portions of the Prodnctns limestone, Salt Mange, and also the 
Mersey beds of Tasmania. 
( c ) Consequently the Karoo Formation, the Damuda-Panchet Series, 
the Ceratite beds of the Salt Mange, and the Hawkesbury-Wianamatta beds 
arc also to be correlated. 
(d) The further correlations higher up follow by themselves. The 
relations will be shown hereafter in a general table. 
* In Afghanistan, however, as we shall see, there are certain beds which Mr. Griesbach correlates with the 
Talchirs, and which overlie the Productus limestones, alternating in their lowest portion with the same. 
11a 104—89 2 A 
