﻿Manchester Memoirs, Vol. xlix. (1905), No . 15 



The argument from association, though of small weight 

 by itself, has considerable value, when critically employed, 

 as supporting more direct lines of evidence. We owe to 

 M. Grand 'Eury, the veteran palaeobotanist of St. Etienne, 

 an extensive investigation of the distribution of Neurop- 

 teridean fronds and associated seeds in the upper and 

 middle coal-measures of France (Grand 'Eury, '04). He 

 is not only led to the conclusion that the Neuropterideae 

 generally bore seeds, and were, to use his words, primitive 

 Cycadinae with the fronds of ferns, but is able, as he 

 believes, to assign definite types of seeds to particular 

 species of the genera Neuropteris, AletJiopteris, Odojitop- 

 teris, Linopteris and others, on the evidence of constant 

 and exclusive association. It thus appears that the 

 numerous unassigned carboniferous seeds of the type of 

 Trigonocarpon or the seed of Neuropteris JieteropJiylla are 

 likely, as investigation proceeds, to find a place among the 

 fern-like genera. 



To bring this part of our subject to a conclusion, all 

 the evidence points to the inference that a large part 

 (probably a decided majority) of the fern-like plants of 

 the carboniferous flora were Spermophyta, allied most 

 nearly to the Cycadacese among recent plants, but retain- 

 ing, in their vegetative structure, clear indications of 

 affinity with the ferns. For this great synthetic group the 

 name Pteridospermese has been proposed.* (Oliver and 

 Scott, '04) 



* Since this lecture was delivered, M. Grand 'Eury has discovered, in 

 the well-known species Pecopleris Pluckeneti, a new and striking example of 

 .a "seed-bearing fern." The specimens, which the discoverer most kindly 

 allowed me to examine, show a multitude of small seeds (a species of 

 Carpolithcs) ranged along the pinnae of the fronds, which scarcely differ from the 

 ordinary vegetation form. In favourable cases the connection between the 

 seeds and the rachis appeared evident. 



The fact that a species referred to the genus Pecopteris should thus 

 prove to be a member of the Pteridospermeai is a surprising one, considering 



