3 o6 SCIENCE PROGRESS 



on the other. The special interest of Petry's results lies in his 

 very reasonable suggestion that Botrychium and Helmintho- 

 stachys have been derived from an ancestor which branched 

 freely, and in the strengthening of the evidence pointing to a 

 close relationship of this isolated living fern family to the most 

 primitive extinct ferns. 



Flowering Plants. — From the extensive literature dealing 

 with the flowering plants we select, as being of evolutionary 

 rather than merely systematic or geographical interest, the 

 following contributions. Two interesting papers, both of the 

 nature of preliminary notes, deal with the morphology of the re- 

 markable genus Gnetum. Pearson (Journ. Linn. Soc. Bot. 43) de- 

 scribes various new features observed in the flower which lead 

 him to believe that the primitive form was hermaphrodite ; 

 that the flower-envelope has arisen by specialisation of a barren 

 leaf structure ; that the antherophore is of leaf origin and has 

 been formed by fusion of two filaments ; and that the primary 

 endosperm of Gnetum is entirely homologous with that in the 

 allied Welwitschia and cannot be regarded as a true prothallus, 

 but is a new structure not found in other plants. To this " new 

 morphological entity, neither sporophytic nor gametophytic," 

 Pearson gives the name " trophophyte " and suggests that the 

 endosperm of Angiosperms may be a highly specialised form 

 of this trophophyte. Thompson (Amer. Journ. Bot. 2) has 

 also studied the embryology of several species of Gnetum, and 

 he too finds evidence pointing to a much closer relationship 

 between Gnetales and Angiosperms than has hitherto been 

 obtained. He states that the pollen-grains may germinate 

 well up in the so-called " style," the pollen-tubes growing 

 down to the nucellus as in Angiosperms ; that no male prothal- 

 lial cells are produced ; that only free nuclei (never cells) are 

 formed in the embryo-sac before the tubes enter (their entry 

 being also immediately preceded by the organisation of one or 

 more eggs) ; that before fertilisation the female gametophyte 

 becomes divided into numerous multinucleate compartments 

 in each of which all the nuclei then fuse ; and that the endo- 

 sperm is formed by divisions of the fusion nuclei in the lowest 

 compartments. The results of these two important papers 

 will certainly figure largely in future discussions regarding the 

 origin of the Angiosperms. A long paper by Miss Sargant and 

 Mrs. Arber (Ann. Bot. 29), deals with the morphology of the 



