478 Palaeontologie. 



Oliver, F. W., Notes on Trigonocarpns Brongn. and 

 Polylophospermum Brongn., two genera of Palaeo- 

 zoic seeds. (New Phytologist. Vol." III. No. 4. p. 96 and 

 a plate. 1904.) 



It is shown that these two genera possess the same fun- 

 damental type of Organisation as Stephanospenmim, a seed 

 recently re-investigated by the same author. The petrifactions 

 described belong to the Willi am so n Collection in the British 

 Museum (Nat. Hist.), and were originally obtained from the 

 Stephanian of Grand'Croix. Their chief interest lies in the 

 structure of the nucellar wall, and in the possible relations of 

 the nucellus to the integument, and in certain indications which 

 point to the existence of an outer fleshy coat to the testa. 



In the small seed Trigonocarpns piisilliis Brongn., the testa 

 is overlaid by two or three layers of quite thinwalled, iso-dia- 

 metric, parenchyma-cells, which very iikely represent but a 

 portion of a more extensive tissue, thus leading to the con- 

 clusion that Trigonocarpns was a drupaceous seed with a hard 

 Shell or sclerotesta^ and an outer sarcotesta. 



In this seed, the nucellus is like that of Stephanospermiim, 

 in possessing a loose flange or collar round the base of the 

 nucellar apical process, and this process probably represents 

 the epidermis that has become detached from the other tissues 

 of the roof of the pollen-chamber. The epidermis, from the 

 level of the pollen-chamber floor, and extending almost to the 

 chalaza, is completely separated from the nucellus, and in places 

 lies in contact with ihe lining of the testa. Thus it may be 

 supposed that the nucellus stood originally free within the testa 

 from the chalaza upwards, as in Lepidocarpon. An alternative 

 view. that the relations were as in recent Cycads, or as in the 

 fossil seed Lagenostoma, is unsupported by observational data 

 and depends on certain theoretical considerations. 



The whole nucellus below its epidermis appears to have 

 been ensheathed in a series of anastomosing Strands of tracheal 

 Clements, and in this respect differs from the structure obtai- 

 ning in Stephanospermum and Aetheotesta. 



Polylophospermum stephanense Brongn. is an elongate 

 prismatic seed, some 15 mm. in length; with a long micropylar 

 beak. The testa expands at the apex and base into cupular 

 emergences in such a way as to produce a sort of false Cham- 

 ber at either end of the seed, Chambers which enclose respec- 

 tively the micropylar beak, and the pedicel or funicle. The 

 cells of the external limiting layer or epidermis of the nucellus 

 are very curiously modified, a large number, especially in the 

 region of the pollen-chamber, projecting as dome-shaped pus- 

 tules, which suggest that an expansion has occurred as a con- 

 sequence of a mucilaginous breakdown of the contents. 



Strands of transfusion tracheids run in the peripheral 

 tissues of the nucellus, which for the most part, are scalariformly 



