part 3] fossil plants fkom the Falkland islands. 325 



Dr. Halle had been preserved, it would have revealed a well- 

 defined midrib. 



Glossopleris Broivniana Brongniart. 



Leaves showing the characters of this species have been recog- 

 nized from several localities, and, as already stated, some of them 

 in the basal part of the lamina agree very closely with the incom- 

 plete specimen identified by Halle as Qangamopteris. None of 

 the recently acquired material shows any features not already 

 noticed by him in the description of the Falkland specimens. 



Dadoxylon BaJceri, sp. nov. (PI. XXII, figs. 19-22 ; text-figs. 

 4&5, p. 327.) 



Several specimens of silicified wood were collected by Dr. Baker 

 at Walker Creek and Fanny Cove, on the southern side of Choiseul 

 Sound, East Falkland. They all exhibit the same characters of 

 the secondary wood, and cannot be specifically separated. The 

 largest piece, assuming that the pith was centric, must have 

 belonged to a stem at least 25 cm. in diameter. Another speci- 

 men, S cm. hy 5 5 cm. in cross-section, shows portions of the 

 medullary and perimedullary regions. The preservation of the 

 smaller specimen is poor, but one can distinguish in the pith and 

 primary medullary raj's large parenchymatous cells, some of which 

 undoubtedly had a secretory function, and resemble those described 

 by the late Miss Holden in Dadoxylon indicum Holden from the 

 Barakar Beds of India. 1 There are no indications of secretory 

 reservoirs or canals like those described by Halle in Dadoxylon 

 lafoniense Halle, 3 a petrified branch from Darwin Harbour, East 

 Falkland. Our specimen resembles the Indian stem in having a 

 zone of tissue lining the inner side of the primary bundles, which 

 differs from the pith and the wood in the shape and size of its 

 constituent elements. In the Indian stem this zone consists of 

 ' transfusion-tracheids '. It is impossible to say whether the Falk- 

 land stem agrees in this respect with D. indicum ; but, in the 

 appearance of the transverse sections, the two forms are strikingly 

 alike. In transverse section the primary bundles at the periphery 

 of the pith form rather more acute wedges than in I), lafoniense, 

 D, indicum, and another species D. Tchihatcheffi (Zalessky) 3 

 from the Permian of Kuznetsk, which resembles our type in some 

 other respects. Warren 4 has described a stem from the Coal 

 Measures of Natal (Upper Ecca) in which irregularly reticulate 

 elements occur internal to the protoxylem. He also figures 5 

 tracheids with pitting very similar to that observed in Dadoxylon 

 Baker i. 



The secondary wood is more clearly preserved in some of the 

 other fragments. In transverse section (fig. 22, PI. XXII) it is 



1 HoMon (17). 4 Warren (12) p. 352. . 



2 Halle (11) pi. ix. ' Warren (12) text-fig. IB, p. 353. 

 11 Zalessky (11). 



