422 BBNNETTITALES [CH. 



which was believed to have borne the flowers for which the 

 new designation was proposed. Two other species, Williamsonia 

 hastula and W. pecten, were assigned to the new genus. The 

 conclusion arrived at by WilUamson as to a connexion between 

 Zamites gigas and Williamsonia flowers was, however, not accepted 

 by Saporta^, who figured and described several exceptionally 

 good specimens from the Yorkshire coast which formed part of 

 the Yates collection in the Paris Natural History Museum. In 

 1897 a short account was published^ of the Yates specimens, an 

 examination of which convinced me of the correctness of William- 

 son's views as to an organic connexion between stems, peduncles, 

 flowers, and fronds. During a visit to Paris several photographs 

 were taken, but these were not published : similar photographs 

 have since been reproduced by Wieland^ and reduced copies from 

 two negatives in my possession are shown in figs. 541, 542. The 

 restoration by Williamson in his weU-known memoir is probably 

 correct so far as the general habit of the plant is concerned, 

 though the flowers which he speaks of as male are now known to 

 be ovulate. The position of the male organs, whether borne 

 separately or on the same axis as the megasporophylls, has not 

 been definitely settled. 



In 1891 the Marquis of Saporta thus introduced his discussion 

 on Williamsonia, — 'avec les Williamsonia nous abordons un des 

 problemes les plus difficiles, un des sujets des plus controverses, 

 mais aussi les plus curieux, peut-etre meme le plus remarquable 

 de tous ceux que nous offre I'ensemble des plantes jurassiques.' 

 Wieland's investigations have placed us in possession of many 

 important facts with regard to the closely allied flowers of Cyca- 

 deoidea and have enabled us considerably to extend our knowledge 

 beyond the stage represented by the work of Carruthers, William- 

 son, and other authors ; and more recently Nathorst's important 

 discoveries have demonstrated the close agreement between 

 Williamsonia and Bennettites (Cycadeoidea). Several problems 

 still remain -unsolved. Having regard to the deficiency of the 

 data concerning the morphology of the Williamsonia type of 

 flower and the wider question as to a phylogenetic relationship 

 that some botanists believe to exist between the Bennettitales 



1 Saporta (91). ^ geward (97^). ' Wieland (11). 



