92 THE EVOLUTION OF PLANTS 



being pinnate are simple, the pollen -sacs are 

 inserted in two rows, directly on the main stalk. 

 This is a distinct step towards the simpler type 

 of stamen, and it is interesting to note that the 

 occurrence of such a stage was predicted theoreti- 

 cally, by Arber and Parkin, two years before its 

 discovery in the Mexican fossil. It may also be 

 mentioned that in Wielandiella, the plant referred 

 to above in speaking of the external characters, 

 the stamens, according to Prof. Nathorst, were 

 comparatively simple structures. 



It is curious that in some respects the older 

 Cycadophytes (Wielandiella and the Mexican 

 WiUiamsonia) seem to approach the Angio- 

 sperms more nearly than those from later Meso- 

 zoic strata. This may perhaps mean that the 

 typical Bennettitese were a side-branch, a little 

 further removed than their predecessors from the 

 main-line of evolution of the Flowering Plants. 



The peculiar structure of the gynseceum 

 in all known Bennettitese separates them at 

 present from the Angiosperms, and as yet no 

 transitional forms have been discovered. The 

 ovules, it will be remembered, are seated each 

 on a long stalk, and between them are the sterile 

 scales, which expand at their ends to form col- 

 lectively the pericarp. The stalk on the end of 

 which the ovule is borne is best regarded as a 

 much simplified carpel; the barren scales may 



