CYCADOPHYTA 597 



in like manner the limits of the Bennettitales are at 

 present identical with those of the Bennettiteae, 

 if we include the Williamsonieae, which should perhaps 

 count as a distinct Order or Sub-order. 



Throughout our description of the Bennettiteae we 

 have assumed that these plants really had affinity 

 with Cycads. The sum of their vegetative characters 

 can leave no doubt that this supposition is justified, 

 but it is a striking fact that in their reproductive 

 organs, that is to say, in the very characters on which 

 systematists are accustomed to rely, the Bennettiteae 

 differ toto coelo from the Cycadaceae. The organisation 

 of the flower is so different from that of the cones of 

 any recent Cycad that it is difficult to determine 

 with any certainty the homologies of the parts in the 

 two groups. It is, of course, possible that future 

 palseontological discoveries may reveal the existence 

 of types of fructification intermediate between those 

 of Cycadales and Bennettitales ; at present, however, 

 it seems as if their development had followed different 

 lines from a very early stage. 



The points in v/hich Bennettitales agree with 

 Cycadales are, the form and general structure of the 

 stem, the histological details of its tissues, the arrange- 

 ment of the leaves, their form and minute structure ; 

 in all these respects the agreement is exact, so 

 far as the existing investigations on American and 

 European species enable us to judge. The only differ- 

 ences worth mentioning, in the vegetative organs, are 

 the simpler course of the leaf-trace bundles in Bennet- 

 titales, and the presence of ramenta like those of Ferns, 

 instead of the woolly hairs of Cycads. In these two 



