The Present Position of Palaeozoic Botany. 209 



The first of these cases was described by Dr. David White in 

 1904. in a plant named by him Aneimites fertilis, from a Lower Car- 

 boniferous horizon in West Virginia. The frond is a liighly com- 

 pound one, of the form familiar under the designation Adiantites, a 

 generic name which has been discarded on technical grounds of 

 nomenclature. The fructification (originally named WarcUa fertilis) is 

 borne on the apices of branched, terminal extensions of the peripheral 

 pinnae, the pinnules being greatly reduced on the adjacent sterile 

 portions of the frond. The small seeds are rhomboidal in form, lenti- 

 cular in cross-section, and winged ; it thus appears that they were of 

 the platyspermic (bilaterally symmetrical) type. The author believed 

 that he detected traces of a micropyle and pollen-chamber, but in the 

 absence of structural specimens such indications are necessarily ob- 

 scure. In spite of the important différences in the seeds, Dr. White 

 is inclined (mainly on the ground of frond-characters) to regard his 

 Aneimites as more closely allied to Lyginodendron than to any other 

 type of Cycadofilices known up to that time. He points out that the 

 discovery of Pteridospermic characters in Aneimites throws serious 

 suspicion on the sterile frond-genus Eremopteris among others. My 

 friends, Mr. Arber and Prof. F. W.Oliver inform me that they 

 liave found strong evidence for the occurrence of seeds, comparable to 

 those of Aneimites, in a species of Eremopteris. 



A few months later, M. Grand'Eury (in April, 1905) made his 

 striking' discovery of the seeds of Pecopteris PlucJîeneti, from the Upper 

 Coal-Measures of St. Etienne. In 20 specimens he found the seeds 

 attached by hundreds to the fronds ; they may occur on the ordinary, 

 unmodified foliage, but where they are numerous the lamina is some- 

 what reduced. The small oval seeds (named Carpolithes granu- 

 latus by Grand'Eury nearly 30 years earlier) are attached to the 

 ends of the principal veins, and are provided with a border or wing; 

 their form is so similar to that of Samaropsis that they may easily 

 be confounded, in the detached condition, with this Cordaitean seed. 

 The resemblance of the seeds of Pecopteris PhicJceneti, Aneimites and 

 (I believe) Eremopteris to those of Cordaiteae is a striking fact, 

 showing that the bilateral or radial symmetry of the seed is of no 

 value as a means of distinction between this Gymnospermous family 

 and the Pteridosperms. 



No palaebotanical discovery would be more interesting at the 

 present time than that of the anatomical structure of Pecopteris 

 Plucl-eneti. The species is not quite a typical one, and was placed 

 by S t e r z e 1 (1886) in a distinct genus, DicJisoniites, on account of his 

 discovery of bodies which he interpreted as sori, and compared to 

 those of Didsonia; their nature is not j'et clearly understood. It 

 seems certain that other fronds of the Pecopteris type will prove to 



Progressus rei botanicae I. 



14 



