282 PEOF. A. C. SEWAKD OX A COLLECTION OF [vol. Ixxviii, 



PlANTA IISX'EET^ SEDIS. (PI. XIII, fig. 10.) 



On one piece of sandstone there are faint impressions of crowded 

 branched filaments, some of which are shown in fig. 10. Their 

 occurrence in a fairly-dense mass reminds us of the branched 

 leaves of Dicranopliyllum Ricliiri Penier ^ from the Coal Measures 

 of Belgium ; but they are more probably portions of pinnules of a 

 fern-like plant, such as some of the Lower Carboniferous species 

 referred to Rhodea or Sphenopteridiiim. Generic determination is 

 hardh^ possible. 



Conclusion. 



Mr. Douglas, in his recent paper on ' Geological Sections through 

 the Andes of Peru & Bolivia,' referred to the importance of 

 obtaining accurate determinations of the plants which he collected 

 from the Peruvian strata.^ After a careful examination of the 

 material, I submitted most of the specimens to Dr. Kidston,^ who 

 kindly expressed his opinion thereon. My first impression led me 

 to regard the Peruvian plants as Upper Carboniferous in age, and 

 this conclusion, if correct, would raise a question of considerable 

 interest, in view of the fact that no member of the Glosso]7teris 

 flora is included in Mr. Douglas's collection. Dr. Kidston, however, 

 is inclined to regard the palseobotanical evidence as more favourable 

 to a Lower Carboniferous horizon. 'A re-examination of the 

 specimens in the light of his suggestions caused me to modify my 

 conclusions. The plants are too imperfect to serve as trustworthy 

 guides : the rocks may belong to the upper part of the Car- 

 boniferous formation or, on the whole more probabl}^, to the Lower 

 Carboniferous. Lower Carboniferous plants similar to Northern 

 Hemisphere types have previously been recorded from South 

 America ; but, so far as I am aware, no L^pper Carboniferous flora 

 is known from South America composed exclusively of Northern 

 Hemisphere tj^pes, with no admixture of members of the Glosso- 

 pteris flora. The inclusion of two Wealden species in the list 

 given by Fuchs must be attributed to incorrect determinations ; 

 none of the specimens collected by Mr. Douglas can reasonably be 

 referred to a Mesozoic flora. Further research is greatly to be 

 desired, since the available data are inadequate as a basis for any 

 positive statement. 



Mv thanks are due to Mr. E. E. Holttum, of St. John's Collesre, 

 Cambridge, for assistance given to me in the preparation of this 

 paper. 



^ Eenier (10) pi. cxvii. 

 2 Doug-las (21) p. 250. 



^ I am greatly indebted to my friend, Dr. Eobert Kidston, F.E.S., for his 

 helpful criticisms and suggestions. 



