578 REPORT— 1880. 



Gaj Miocene. 



1^ . \ Upper , . Lower miocene (?). 



^^^ * ' ] Lower . . Oligocene. 



Khirthar. . j YZ7v '. 1 Eocene. 

 Ranikot J 



There is little doubt that the upper Manchhars, 5000 feet thick, correspond to 

 upper and middle Siwalik. 



The Gaj marine fauna is clearly Miocene, and rather upper than lower. My 

 original view of the age, hased chiefly on the Echinodermata, has heen confirmed 

 hy Professor Martin Duncan's examination of the corals. A considerable number 

 of mammalia have been foimd in the lower Manchhar beds, which pass downwards 

 into the Gaj, and consequently cannot be older than upper Miocene. 



Turning now to the Siwalik mammalian fauna, it is, I believe, with the addition 

 of the species lately described by Riitimeyer and Bose, composed of 48 genera, com- 

 prising about 93 species, 23 genera being extinct, and 25 recent. 



Amongst the fossil genera, 12 are imknown elsewhere ; 4 — Pseudalurus, Amphi~ 

 cyan (the occurrence of tbis in true Siwalik beds is doubtful), Listriodon and Dor- 

 catherium are not known in beds later than Miocene in Europe, {Pseudalurus is 

 found m Pliocene beds in America) ; 7 are Miocene and Pliocene of the recent genera ; 

 9 range back in Europe to upper Miocene ; 10 only to Pliocene ; and 6 are only 

 known elsewhere as li\'ing genera, or are found in Post -pliocene deposits. The very 

 large proportion of species belonging to recent genera, like Felts, Canis, Ursus, 

 Elephas {Euelephas and Lo.vodon), Equus, Cervus, Bos, Antilojte and Capi-a, gives a 

 singularly late aspect to the fauna. 



Now, in the lower Manchhars of Sind no recent genera have been found except 

 Jihinoceros, Sus and Manis (the generic identification of the latter being excessively 

 doubtful), whilst Dinotherium, Antliracotlierium, Hyopotmnus, Hyotherimn, and' 

 two new genera allied to Merycopotmnus are found, none of which occur in the 

 Siwaliks proper, and the species common to the two, excluding Amphicyon, are 

 8 in number, consisting of forms of Mastodon, RJiinoceros, Acerotherium, Sus, Chali- 

 cotherium, and Dorcotherium. There are 10 extinct genera to two recent {Manis 

 beino- omitted as doubtful). There can be no question that the Manchhar fauna is 

 decidedly older than Siwalik. But Manchhar cannot be older than upper Miocene, 

 therefore Siwalik is Pliocene. 



This conclusion is supported by the facts that out of six forms of reptiles sufli- 

 ciently known to afford means of comparison, three are recent species ; that amongst 

 the four or five kinds of birds hitherto determined, one is probably identical with 

 the livins ostrich ; and that all the land and fresh-water shells found are forms 

 still existing. No such connexion with the recent fauna is known in any true 

 Miocene rocks. The geological evidence is also in favour of a newer age for the 

 fossiliferous Siwalik beds, as they form the upper members of the series ; whilst 

 the lower Manchhar rocks are at the base of formations whicli apparently represent 

 the Siwalik series. 



The sino-ularly Miocene aspect of a portion of the Siwalik mammalian fauna finds 

 a parallel in the case of Pikermi, in Attica ; although the age of the one does not 

 necessarily determine that of the other. The superb collection of remains made 

 and admirably described by Professor Gaudry has constantly been classed as Mocene 

 bv the describer. It comprises 30 genera, of which 19 are extinct. Of these 30 

 o-'enera, 1 3, besides HeUadofherium, which is very possibly identical with Sivatherium, 

 are found in the Siwaliks of India. The ruminants are in both very much more 

 numerous in proportion to the other artiodactyle ungulates than is usual in Miocene 

 strata • in the Siwalik, 33 against 10 ; in the Pikermi beds, 15 ruminants to 1 pig 

 and 1 ^Chalicotherium. Nevertheless, out of 42 species found at Pikermi, no less 

 than 15 are common to Miocene deposits in other parts of Europe. The connexion 

 between Pikermi beds and typical Miocene is thus very similar to that between 

 Siwalik and Manchhar beds. 



But the Pikermi beds at their base contain three characteristically Pliocene 



