442 



NA TURE 



[Sept, 4, 1884 



so many typically Middle Tertiary forms, such as Dinotherium, 

 Anthracotkerium, and Hyopotamus, shows a great change. The 

 Mollusca tell the same tale. All the forms known from the 

 upper Siwaliks, with one exception, are recent species of land 

 and fresh-water shells now living in the area. Of seven fresh- 

 water Mollusca (Mem. G.vl. Surv. Ind. vol. xx. pt. 2, p. 129) 

 found associated with the lower Siwaliks none appears to be 

 identical with any living species, and only two are allied, one 

 closely, the other more remotely, to forms now met with in 

 Burmah, 30 of longitude further east. 



Before proceeding with the argument, it is as well to call 

 attention to the very important fact just mentioned. It has been 

 asserted over and over again that species of Mammalia are 

 peculiarly short lived, far more so than those of Mollusca. In this 

 case, so far as the evidence extends at present, one-third of the 

 species of Mammalia survived the changes that took place, 

 whereas not a single mollusk is found both in the upper and 

 lower Siwaliks. It should be remembered that the recent 

 molluscan river fauna of this part of India is very poor in 

 species, and that we probably know a considerable proportion of 

 that existing in Siwalik times. 



The geological age of the lower Siwalik beds of Sind is 

 shown by their passing downwards into marine fossiliferous beds, 

 known as the Gaj group, of Miocene age, the following being 

 the section of Tertiary strata exposed in the hills west of the 

 Indus : — 



( Upper 

 •- I Lower 



Gaj. 

 Nam 



5000 unfossiliferous 



3:100 to 5000 fossiliferous . 



1000 n 1500 fossiliferous . 

 4030 t) 6003 unfossiliferous 



100 to 1500 fossiliferous . 



500 to 3000 fossiliferous . 



6nijj fossilifur ius . . . 



Oligocene 

 ■ Eocene 



Clearly the lower Siwaliks of Sind cannot be older than 

 Upper Miocene ; therefore the upper Siwaliks, which are shown 

 by both biological and geological evidence to be of much later 

 date, must be Pliocene. 



Gondwdna System of India. — In the peninsula of India there 

 is a remarkable deficiency of marine formations. Except in the 

 neighbourhood of the coast or of the Indus Valley there is, with 

 one exception (some Cretaceous rocks in the Nerbudda Valley), 

 not a single marine deposit known south of the great Gangetic 

 plain. But in Bengal and Central India, over extensive tracts 

 of country, a great sequence of fresh-water beds, probably of 

 fluviatile origin, is found, to which the name of Gondwana 

 System has been applied. The uppermost beds of this system, 

 in Cutch to the westward, and near the mouth of the Godavari 

 to the eastward, are interstratified with marine beds containing 

 fossils of the highest Jurassic (Portlandian and Tithonian) 

 types. 



The Gondwana system is a true system in the sen^e that all 

 the series comprised are closely connected with each other by 

 both biological and physical characters, but it represents in all 

 probability a much longer period of geological time than do any 

 of the typical European systems. The highest members, as 

 already stated, are interstratified with marine beds containing 

 uppermost Jurassic fossils. The age of the lowest members 

 is less definitely determined, and has been by different writers 

 classed in various series from Middle Carboniferous to Middle 

 Jurassic. The Gondwana beds from top to bottom are of 

 unusual interest on account of the extraordinary conflict of 

 palxontological evidence that they present. 



The subdivisions of the Gondwana system are numerous, and 

 in the upper portions especially the series and stages are different 

 in almost every tract where the rocks are found. The following 

 are the subdivisions of most importance on account of their 

 fauna and flora, or of their geological relations : — 



( Cutch and Jabalpur 

 Upper Gondwana ... J Kota-Maleri 

 ( Rajmahal 



SPanchet 

 "*■ ■ • -{S J:mdKdmthi 

 f Karharbari 

 I Talchir 



The upper Gondwanas, where best developed, attain a thick- 

 ness of 11,000 feit, and the lower of 13,000 ft. 



The Talchir and Barakar subdivisions are far more generally 

 present than any of the others. 



Talchir. — The Talchir beds consist of fine silty shales and fine 

 soft sandstone. Very few fossils have been found in them, and 



these few recur almost without exception in the Karharbari 

 stage. The Talchirs are principally remarkable for the fre- 

 quent occurrence of large boulders, chiefly of metamorphic 

 rocks. These boulders are sometimes of large size, 6 feet or 

 more across, 3 to 4 feet being a common diameter; all are 

 rounded, and they are generally embedded in fine silt. 



Karharbari. — The Karharbari beds are found in but few 

 localities. They contain some coal-seams, and the following 

 plants have been met with (Feistmantel, Palceontologia Indica, 

 ser. xii. vol. iii.) : — 



Conifer.*. — Euryphyllum, I sp. ; Voltzia, 1 ; Albertia, I ; 



Samaropsis, I. 

 Cycadeace.e. — Glossozamites, 1 ; Noeggeralhiopsis, 1. 

 FlLICES. — Neuropteris, I ; Glossopteris, 4 ; Gangamopteris, 4 ; 



Sagenopteris, 1. 

 Equisetacea. — Schizoneura, 2 ; Vertebraria, 1. 



The most abundant form is a Gangamoptcris. The Voltzia 

 (V. heterophylla) is a characteristic Lower Triassic (Bunter) 

 form in Europe. The Neuropteris and Albertia are also nearly 

 related to Lower Triassic forms. The species of Gangamoptcris, 

 Glossopteris, Vertebraria, and Noeggeralhiopsis are allied to 

 forms found in Australian strata. 



Damuda. — The Damuda series consists of sandstones and 

 shales with coal-beds ; the floras of the different subdivisions 

 present but few differences, and the following is the list of plants 

 found 'Pal. Ind. ser. ii. xi. xii. vol. iii.) : — 



Conifer.*. — Rhipidopsis, 1 sp. ; Voltzia, 1 ; Samaropsis, 1 ; 



Cyclopityi, I. 

 Cycadeace.e. — Pterophyllum, 2; Anomozamites, 1; Noeg- 



gerathiopsis, 3. 

 FlLICES. — Sphenopteris, I; Dicksouia, I; Alethopteris, 4; 



Pecopteris, I ; Meriauopteris, I ; MacroUt niopteris, 2 ; 



Palceovittaria, I ; Angiopteridium, 2 ; Glossopteris, 19 ; 



Gangamoptcris, 7 ; Belcmnopteris, 1 ; Anthrophyopsis, 1 ; 



Dictyopteridiitm, I; Sigenopteris, 4; Actinopteris, I. 

 Equisetace.-E. — Schizoneura, I ; Phyllotheca, 3 ; Trizygia, I ; 



Vertebraria, I. 



The only remains of animals hitherto recorded are an Esthcria 

 and two Labyrinthodonts, Brachyops laticcps and an undescribed 

 form formerly referred to Archegjsaurus. The only European 

 genus allied to Brachyops is of Oolitic age. 



The most abundant of the above-named fossils are Glossopteris 

 and Vertebraria. With the exception of Nofgeerathiopsis all the 

 cycads and conifers are of excessive rarity. More than one-half 

 of the species known are ferns with simple undivided fronds and 

 anastomosing venation. 



For many years European palaeontologists generally classed 

 this flora as Jurassic. 1 This was the view accepted by De Zigno 

 and Schimper, and, though with more hesitation, by Bunbury. 

 The species of Phyllotheca, Alethopteris (or Pecopteris), and 

 Glossopteris (allied to Sagenopteris) were considered to exhibit 

 marked Jurassic affinities. It was generally admitted that the 

 Damuda flora resembles that of the Australian Coal-Measures (to 

 which I shall refer presently) more than it does that from any 

 known European formation ; but the Australian plants were 

 also classed as Jurassic. There is no reason for supposing that 

 the more recent discoveries of Damuda plants would have modi- 

 fied this view ; the identification of such forms as true Sagenopteris 

 and the cycads Pterophyllum and Anomozamites would assuredly 

 have been held to confirm the Jurassic age of the beds. So far 

 as European fossil plants are concerned, the Damuda flora 

 resembles that of the Middle or Lower Jurassics more than any 

 other. 



One form, it is true, the Schizoneura, is closely allied to 

 S. paradoxa from the Bunter or Lower Trias of Europe. Other 

 plants have Rhcetic affinities. But the connections with the 

 Triassic flora do not seem nearly equal to those shown with 

 Jurassic plants, and the reason that the Damuda flora has been 

 classed as probably Triassic must be sought in the impossibility 

 of considering it newer ( Feistmantel, Pal. Ind. ser. xii. vol. iii. 

 pp. 57. 129, &c), if the next overlying stage is classed as 

 Upper Trias or Rhretic, and in the close affinity with the under- 

 lying Karharbari beds, which contain several Lower Triassic 

 types. 



Panchet. — The uppermost series of the lower Gondwanas con- 

 sists chiefly of sandstone, and fossils are rare. The most in- 



1 De Zigno, Flora FossilU Form. Got. pp. 50, 53 ; Schimper, Traiti* dc 

 PalimtologU vigttale, i. P645 ; Bunbury, Q. J . G. o". 1861, xvii. p. 350. 



