Sept. 4, 1884] 



NA TURE 



441 



Prof. Gaudry evidently has no very profound faith in this 

 hypothesis, and it is unnecessary to refute it at length. One fact 

 is sufficient to show that it is untenable. However sudden may 

 have been the cataclysm that is supposed to have broken up the 

 Miocene plains of Attica, a very long period, measured in years, 

 must have elapsed before the Pliocene marine fauna could have 

 established itself. Now, the bones of mammals exposed on the 

 surface decay rapidly ; the teeth break up, the bones become 

 brittle. It is doubtful if bones that had been exposed for only 

 five or six years would be washed down by a stream without 

 being broken into fragments ; the teeth especially would split to 

 pieces. The condition of the Pikermi fossils proves, I think, 

 that they must have been buried very soon after the animals died, 

 that they were not exposed on the surface for any length of time, 

 and that they could not have been washed out of an earlier 

 formation, and it appears to me incredible that the Pikermi 

 mammals were not contemporary with the Pliocene Mollusca that 

 occur in the same beds. In short, I cannot but conclude that the 

 Pikermi mammals were Pliocene and not Miocene. 



This view is entirely in accordance with the opinions of 

 Theodor Fuchs (Denkschr. K. Acad. IViss. Wien, 1877, xxxvii. 

 2 C Abth. p. 1). He has given a good account of the geology of 

 various places in Greece, and amongst others of Pikermi. He 

 found, again, the conglomerate with Pliocene marine Mollusca 

 interstratified with the basal portion of the mammaliferous beds, 

 and he concludes (I.e. p. 30), that not only is it clear that these 

 mammaliferous beds arc of Pliocene age, but that a comparison 

 of their geological position with that of the marine strata of the 

 Piraeus proves that the Pikermi beds occupy a very high position 

 in the Pliocene, and are probably the highest portion of the 

 system as developed in the neighbourhood. 



Fuchs also shows that the principal Pliocene mammaliferous 

 beds are of later date than the typical Pliocene (sub-Apennine) 

 beds of Italy, and that some Mammalia found associated with the 

 latter comprise forms identical with those of the Pikermi beds. 

 In subsequent papers on the age of the beds containing Hipparion 

 the same writer shows reasons for classing these strata in Italy, 

 France (Vaucluse), and Germany as intermediate between 

 Miocene and Pliocene. This leaves the difficulty unsolved, for 

 he had shown the Pikermi beds to be high in the Pliocene 

 system. They rest unconformably upon certain fresh-water 

 limestones, clays, &c, containing plants and Mollusca, and 

 classed by Gaudry as Miocene, but by Fuchs as Pliocene. Thus 

 by both writers mammaliferous beds of Pikermi are referred to 

 a considerably later geological horizon than those containing 

 identical species in other parts of Europe. 



It would require too much time to enter into the still more 

 difficult question of the various plant-bearing beds in different 

 parts of Europe and in Greenland containing a flora classed by 

 Heer and others as Miocene. Gardner has given reasons for 

 considering the Greenland beds Eocene ; Fuchs, as just stated, 

 is of opinion that the Greek beds are Pliocene. One point should 

 be noted, that the more nothern flora is considered older than the 

 more southern, and it will be remarked that the same observation 

 applies to the supposed Upper Miocene fauna of France and 

 Germany and the Pikermi fauna of Greece. 



i. — The next instance which I shall describe is another 

 of the most important fossil mammalian faunas of the Old World, 

 that found in the Upper Tertiary beds that fringe the Himalayas 

 on the south. The name applied to this fauna is taken from one 

 of the localities in which it was first found, the Siwalik (correctly, 

 I believe, Shib-wala) hills, between the Deyra Dun and the 

 plains north by east of Delhi. Bones of Siwalik Mammalia are 

 found, however, throughout a considerable area of the Northern 

 Punjab. 



The Siwalik fauna has been worked out, chiefly by Falconer 

 and Lydekker, the last-named being still engaged in describing 

 the species. The following is a list of the genera found in the 

 true Siwalik beds : — 1 



MAMMALIA. 



PRIMATES.— Pal&opitheeus, 1 sp.; Macaais, 2; Semnopilliecus, 



1 ; Cynopithecus, 2. 



Carnivora. — Mustela, 1; Mellivora, 2; Mellivorodon, 1; 

 Lutra, 3 ; Hy<enodon, 1 ; TTrsus, 1 ; Hycenardus, 3 ; Canis, 



2 : Viverra, 2: Hyana, 5; Lepthyana, 1 ; sEluropsis, I; 

 /Elurogale, 1 ; Felts, 5 ; Macharodus, 2. 



Proboscidea.— Elephas, 6 (Euelephas, I ; Loxodon, 1 ; 

 Stegodon, 4) ; Mastodon, 5. 



1 Lydekker./. AS. B. 18S0, pt. 2, p. 34; Paleumtolog'ut Indica, si r. x 

 vols. i. ii. lii ; Records Geo!, Sins'- India, 18S3, p. 81. I am indebted to 

 Mr. Lydekker for some unpublished additions. 



Ungulata. — Chalieotherium, 1 ; Rhinoceros, 3 ; Equw, I ; 



Hipparion, 2 ; Hippopotamus, 1 ; Tetraconodon, I ; Sus, 5 ; 



Hippohyus, 1 ; Sanitkerium, i ; Merycopotamus, I ; Cervus, 



3; Dorcatherium, 2; Tragulus, 1; Prnpalaomeryx, I; 



Camelopardalis, 1 ; Helladotherium, 1 ; Hydaspitherium, 2 ; 



Sivatherium, 1 ; Alcephalus, 1 ; Gazella, 1 ; Antilope, 2 ; 



Oreas (?), 1; Palaoryx (?), 1; Portax, 1 ; Hemibos, 3; 



Leptoijs, 1 ; Buhalus, 2 ; Bison, 1 ; Bos, 3 ; Bucapra, 1 ; 



Capra, 2; Ovis, I ; Came/its, I. 

 Rodentia. — Mus, 1 ; Rhyzomys, 1 ; Hystrix, 1 ; Lepus, 1. 



AVES. 

 Graulus, 1; Pelecanus, 2; Megaloscelornis, I ; Argala, 1 ; 

 Struthio, I ; Dromasus, i. 



REPTII.IA. 



CROCOD1L1A. — Crocodilus, I ; Gharialis, 3. 

 Lacertilia. — Varanus, 1. 



Chelonia. — Colossochelys, 1 ; Testudo, 1 ; Bellia, 2 ; Damo- 

 nia, I ; Emys, 1 ; Caulleya, 1 ; Pangshura, I ; Emyda, I ; 

 Trionyx, I. 



PISCES. 

 Bagarius, I. 

 Now, until the last few years, this fauna was classed as 

 Miocene by European palaeontologists as unhesitatingly as the 

 Pikermi fauna still is, and in the majority of European geological 

 works, despite the unanimous opinion of all the geologists who 

 are acquainted with the sub-Himalayan beds, the Siwalik fauna 

 is still called Miocene. The geologists of the Indian Survey, 

 however, class the fossiliferous Siwaliks as Pliocene, on both 

 geological and biological grounds. With regard to the latter, 

 not only does the fauna comprise a large number of existing 

 genera of mammals, such as Macacus, Semnopitueeits, Ursus, 

 Elephas {Euelephas), Equus, Hippopotamus. Camelopardalis, 

 Bos, Hystrix, A/us, and especially Mellivora, Meles, Capra, 

 Ovis, Camelus, and Rhizomys, but three out of six or seven 

 clearly-determined species of reptiles, viz. — Crocodilus palustris, 

 Gliarialis gangeticus, and Pangshura latum — are living forms 

 now inhabiting Northern India, whilst all the known land and 

 fresh water Mollusca, with one possible exception, are recent 

 species. 



These data, however, although very important and very 

 cogent, belong to a class of facts that have led, I believe, in 

 other cases to erroneous conclusions. The geological evidence 

 is far more satisfactory, and it is not liable to the same objection. 

 The whole Siwalik fauna, as given above, has been obtained 

 from the upper beds of a great sequence or system. Beneath 

 the fossiliferous strata at the base of the North- West Himalaya 

 there is an immense thickness, amounting in places to 1. any 

 thousands of feet, of sandstones, clays, and other beds, from 

 none of which recognisable fossils have been procured. The 

 first beds of Known age that are met with below the mam- 

 maliferous Siwaliks are marine rocks belonging to the Eocene 

 system. 



But as we pass from the Himalay-is to the south-vest, along 

 the western frontier of India in the Punjab, and onwards to the 

 south in Sind, the same Siwalik system can be traced almost 

 without in'erruption, and in the last-named country the lower 

 unfossiliferous strata become intercalaled with fossiliferous beds. 

 In Sind the upper Siwaliks no longer yield any vertebrate 

 remains that can be identified, but far below the horizon of die 

 Siwalik fauna a few bones have been found, and the following 

 mammals have been identified (Pal. Pud. ser. x. ; Rec. Geol. 

 Sun: Iud. 1883, pp. 82, &c.)— 



Carnivora. — Amphieyon palaindicus. 



Proboscidea. — Mastodon latidens, M. perimensis, M. fal- 

 coneri. M. pandionis, M. angustidens, Dinotherium indi- 

 iiim, I). sindiense, D. pentepotamim. 

 Ungulata. — Rhinoceros sivalensis, var. inlerm Juts, Acero- 

 theriumperimense, A. blanfordi, Sushysudri us, ffyotherittm 

 sindiense, Anthracotherium silistrense, A. hyopotamoides, 

 Hyopotamus palaindicus, H. giganteus, Hemimeryx blanfordi 

 Sivameryx sindiensis, Agriocharus 'p., Dorcatherium majus, 

 D. minus. 

 Edentata. — Main's (?) sindiensis. 



Although about one-third of the species above named have 

 been found also in the upper Siwalik beds of the Punjab, it is 

 unnecessary to point out in detail why the lower Siwalik fauna 

 is clearly by far the older of the two. The absence of such 

 living genera as Elephas, Bos, Equus, &c, and the presence of 



