«60 



SCIENCE 



[N. S. Vol. XXXVIII. No. 984 



and Snow Hill Islands. Gothan, who has de- 

 scribed the fossil woods, has differentiated six 

 forms, all new. Five of these are given spe- 

 cific names and all of the determinable forms 

 are from Seymour Island. They are as fol- 

 lows: Phyllocladoxylon antarcticum, Podo- 

 carpoxylon aparenchymatosum, Dadoxylon 

 {Araucaria) pseudoparenchymatosum, Lauri- 

 noxylon uniseriatum, Laurinoxylon? sp., 

 N othofagoxylon scalariforme. As I have al- 

 ready mentioned, there is, unfortunately, 

 some uncertainty as to their exact age. Part 

 of the specimens representing the Phylloclad- 

 oxylon are Tertiary and the balance are 

 Upper Cretaceous or Tertiary. The Podo- 

 carpoxylon is given as Tertiary and the bal- 

 ance may be either Upper Cretaceous or Ter- 

 tiary. In either case they show that types 

 now regarded as South American or Austral- 

 asian were much more wide-spread in the early 

 'Tertiary or late Cretaceous. It is of some 

 interest to find structural remains of Arau- 

 '-cariese, Lauracese and Nothofagus, since these 

 ■fthree types are also represented in the leaf 

 impressions studied by Dusen. 



The northeastern portion of Seymour Is- 

 land is made up of Tertiary beds. These are 

 mostly marine calcareous sandstones, but with 

 some tuffs containing augite-porphyrite. In 

 these sandstones Nordenskiold discovered leaf 

 impressions which Nathorst reported upon in 

 his brief paper of 1904.^'' They have been 

 monographed by Dusen.^^ The material is 

 abundant but very fragmentary. Dusen rec- 

 ognizes 8Y different forms, of which only 25 

 receive specific names. Both the results and 

 their method of presentation are open to criti- 

 cism. While Dusen has brought to the work 

 an extensive acquaintance with the existing 

 flora of South America, it does not appear 

 that he has an equal knowledge of paleobotan- 

 ical literature and there is a tendency to see 

 an undue resemblance to the existing flora he 

 seems to know best. 



There are 26 different Phyllites sp., some of 

 which are Angiosperms and some Gymno- 

 rsperms. Of the 37 different ferns only nine 



18 Comptes rendus, loo. cit. 



17 Lief 3, 1908. 



are identified and we are treated to the abom- 

 inable array of 10 Sphcenopteris sp. and 18 

 Pecopteris sp., both form-genera that should 

 really be reserved for Paleozoic fern-like re- 

 mains, Sphenopteris being partly, and pre- 

 sumably wholly, Pteridospermic and Pecop- 

 teris being filicalean. With the exception of a 

 Fagus previously described by Dusen from 

 the Straits of Magellan and a Nothofagus de- 

 scribed by Engelhardt from the same region, 

 all of the named species are new to science. 

 They include forms in the following genera: 

 Miconiiphyllum, Lauriphyllum, Mollinedia, 

 Araucaria, Polypodium, Asplenium, Alsophila, 

 Dryopteris, Caldcluvia, Laurelia, Drimys, 

 Lomatia, Knightia, Fagus, Nothofagus and 

 Myrica. 



The first eight of these have their closest 

 affinities with forms in the existing subtrop- 

 ical flora of southern Brazil, while the bal- 

 ance resemble existing species of West Pata- 

 gonia and southern Chili. Dusen concludes 

 that this mixed character is due to differences 

 in altitude at which the Seymour Island 

 plants grew. This may well be the case, but on 

 the other hand the author is apparently un- 

 aware of the polar extension of more equato- 

 rial climates with a mixing of types since as- 

 sociated with temperate or tropical conditions 

 that occurs in the early Tertiary, or to the 

 general lack of well-defined climatic zones in 

 the history of the earth throughout geological 

 times. Many attempts have been made to 

 emphasize the fact that climates like that of 

 the present or the Pleistocene, of which the 

 present is really a part, or of Glossopteris 

 time, or of earlier glacial periods, were the 

 exception and not the rule when all geological 

 time is considered. The consequent lack of 

 extreme cold in the Tertiary when accom- 

 panied by suifieiently humid conditions would 

 answer for the Seymour Island Tertiary flora 

 equally as well as an altitudinal zonation. 



According to Dusen this flora is typically 

 South American, with only slight relationships 

 to the flora of New Zealand (cf. Laurelia) 

 and Australia (cf. Knightia). This is per- 

 haps what would be expected since both tec- 

 tonically and petrographically Graham Land 



