Origin of the AngiosiJerinons Flora of South Africa. 343 



rocks on the coast of Madagascar, South Africa, and India, as well 

 as rocks of still later periods. This fact, together with the existence 

 of deep water (2,000 fathoms) round the north and east of Mada- 

 gascar and round South India without any marked tailing off of the 

 land-slope to its depth, points to the subsidence having necessarily 

 been local, and having ceased to operate peculiarly abruptly at either 

 end." 



From these remarks I would only conclude that the Southern 

 Ocean had already in Cretaceous times made serious inroads on the 

 supposed ancient Indo-oceanic continent w^ithout having necessarily 

 destroyed the main portion of it. This view seems to have al>o 

 been taken by Mr. E. Lydekker, who took part in the discussion 

 which followed the reading of Mr. Gardiner's paper. This view 

 is further strengthened by the fact that while Lower Cretaceous 

 deposits are only known in South Africa as the so-called Uitenhage- 

 beds in the Sundays Eiver valley, Upper Cretaceous deposits are 

 known from various localities as we go further east and north from 

 the neighbourhood of the St. John's Eiver to Madagascar. The 

 whole difficulty of the distribution of the Cape types seems thus to 

 be capable of solution if we go back to early Cretaceous times, and 

 as remains of Angiosperms have been found in other parts of the 

 world in deposits of early Cretaceous age, we seem to be on safe 

 ground in assuming that they also existed in South Africa during 

 that period, but, as I indicated before, I am inclined to go even a 

 step further. As undoubted Angiosperms have been found in early 

 Cretaceous deposits, their roots probably go back into Jurassic times, 

 and I am in hopes that they will be found yet on the Southern 

 Hemisphere in Jurassic deposits, if they have not been found there 

 already. Nobody who has followed recent advances in Palteo- 

 botany will be prepared to say that this is impossible, for the extent 

 to which the clock has been put back recently in Vegetable Palaeon- 

 tology forms one of the most striking results in modern botanical 

 research. Most competent Palaeo-botanists are, w^ith Zeiller, of 

 opinion (quoted by Scott in Progressus Eei Botanicse, i. (1807), 

 p. 177, where the literature is cited) that the ferns, though they 

 were not entirely absent, occupied an altogether subordinate position 

 in the Palaeozoic period. The forms hitherto taken for ferns being 

 Pteridosperms (Cycadofilices). Scott himself, while agreeing as to 

 the age of the Pteridosperms, thinks this statement too sweeping. 

 The Angiosperms, which concern us here chiefly, have been found in 

 the Cretaceous of various countries, but at this period we find already 

 a number of types developed which live to the present day, and 

 there is not only a possibility, but I think a probability, that they 



