part 3] fossil plants from the Falkland islands. 



327 



Fig. 4. — Radial longitu- 

 dinal section of a me- 

 dullary ray of Dado- 

 xylon Bakeri, sp. nov. 

 X 210. 



[The large pits in the field are 

 probably simple pits on the 

 wall of the underlying tra- 

 cheids (see fig. 5).] 



Fig. 5. — Radial section of 

 Dadoxylon Bakeri, shotv- 

 ing the pitting of the 

 medullary - ray cells 

 above the summer-wood. 

 X 210. 



of VIjjl to a maximum of 29 fi; the 

 tracheids flanking the rays have a 

 breadth of 25 /x and 33 /a respectively. 



There is a well-marked group of 

 Dadoxyla characteristic of the Glosso- 

 pteris flora in which there are clearly- 

 defined features differentiating them 

 from the typical Cordaiteae. In the 

 wood of this southern type there is 

 no trace of a discoid pith ; there is a 

 tendency to a uniseriate and often 

 distant arrangement of the tracheidal 

 pits ; large simple pits are found in 

 the ' field ' ; well defined growth-zones 

 often occur, and in some stems 

 secretory cells or canals are present in 

 the pith. In view of the uncertainty 

 of the geological horizon of some 

 of the southern specimens a tabular 

 representation (p. 328) of the anatom- 

 ical characters loses some of its signifi- 

 cance ; but it shows the occurrence of 

 certain features common to a group 

 of woods, included in the compre- 

 hensive and often imperfectly-defined 

 genus Dadoxylon, found in associa- 

 tion with Glossopteris. 



The presence of well - marked 

 growth- zones in these southern stems 

 has been correlated with climatic 

 conditions associated with the glaci- 

 ation demonstrated by the widespread 

 and thick beds of finite in Gond- 

 wanaland. It is interesting to note 

 that Zalessky records the occurrence 

 of wood similar in type to those 

 enumerated associated with members 

 of the Qlossopteris flora in the 

 Petchora district, although there no 

 glacial deposits are known. 



Dadoxylon Bakeri, sp. nov. 1 



Founded on pieces of large stems. 

 Pith parenchymatous, with a narrow 



1 While recognizing the possibility that 

 the wood named after Dr. Baker may be 

 identical specifically with Dadoxylon lafoni- 

 ense Halle, additional characters shown by 

 our material render advisable a distinctive 

 designation. 



2a 



