478 



SCIENCE 



[N. S. Vol. XXXIV. No. 876 



sociation brings about. If we follow with 

 Schroeter the gradual development of a 

 lacustrine vegetation from the reed-swamp 

 through the marsh (or Flaehmoor) to a 

 peat-moor (Hochmoor), we see how one 

 plant association makes place in its turn 

 for another. JMay not the mixture of vari- 

 ous types of vegetation which we meet with 

 in the petrifactions of our coal-seam repre- 

 sent the transition from the open Cala- 

 mitean or Lepidodendroid swamp to a fen 

 or marsh with plentiful peat-formation, 

 due to the gradual filling up of the stag- 

 nant water with plant-remains? Thus in 

 places, at anj^ rate, a transition from aqua- 

 tie to more terrestrial types of vegetation 

 would take place, while the tree-like forms 

 rooted in the deeper water would continue 

 to flourish. The coal-measure swamp in this 

 stage would differ from the tropical swamp 

 of Kooders by a more abundant under- 

 growth of herbaceous and climbing plants, 

 rooted in damp humus and passing off 

 gradually into drier peat. Such an under- 

 growth of Cryptogamie types, mainly 

 Filieinean or Pteridospermic, would have 

 admirable conditions for luxuriant de- 

 velopment, apart from the provision of a 

 suitable substratum for its roots, owing to 

 the narrow xerophytic nature of the foliage 

 on the canopy of the trees under which it 

 grew. 



Here, too, we see the explanation of the 

 striking difference between the micro- 

 phyllous and arborescent Calamites and 

 Lepidodendracefe, and the large ombro- 

 phile foliage of the FilicineiB and Pterido- 

 sperms, which spread out their shade- 

 leaves under the cover of marsh xerophytes, 

 in exactly the same way as Professor Yapp 

 has so admirably depicted for recent plants 

 in his account of the "Stratification in the 

 Vegetation of a Marsh." 



The development of a mesophj'tic vege- 

 tation in the shelter of the marsh xero- 



phytes makes it unnecessary to postulate an 

 obscuration of the intense sunlight by 

 vapors, as was done by Unger and 

 Saporta for the Carboniferous period. 

 The assumption of a variety of conditions 

 of plant life within the same area helps 

 materially to clear up the difficulties pre- 

 sented by the somewhat incongruous oc- 

 currences met vath in the petrified plant- 

 remains. The presence of fragments of 

 Cordaites, mixed with those of Calamites 

 and Lepidodendra, in the coal-balls can 

 not always be explained either by a drift 

 theory, or by conceiving the fragments to 

 be wind-borne ; but, given an area of retro- 

 gressive peat above the ordinary water- 

 level, and even so xerophytic a plant as 

 Cordaites might well establish itself there, 

 its mycorhiza-eontaining roots being well 

 adapted for gro'^vth in drier peat. The 

 curious occurrence of more or less concen- 

 tric rings in the secondary wood of the 

 stem and roots of Cordaites may represent 

 a response probably not to annual varia- 

 tions of climate, but to abnormal periods 

 of drought, which would affect the upper- 

 peat laj^ers, but not the water-logged soil 

 in which were rooted the Calamites and 

 Lepidodendra. 



If, as I suspect, we had in the peat de- 

 posit of the coal-seam a succession of asso- 

 ciations, we ought to find its growth and 

 history recorded by the sequence of the 

 plant-remains, very much as Mr. Lewis has 

 discovered with such signal success in our 

 Scottish peat-bogs. That some differences 

 occur in the plant-remains building up a 

 seam can be noted by a microscopic exami- 

 nation of the coal itself, in which, as Mr. 

 Lomax tells me, the spores of Lepidoden- 

 dra occur in definite bands. But no system- 

 atic attempt has as yet been made to in- 

 vestigate from this point of view the seams 

 charged with petrified plant debris. Be- 

 fore the Shore pit, which was reopened last 



