i64 THE LIVING CYCADS 



Dioon second, and Zamia is the final genus. There is no 

 difficulty in reading the series, for a sporophyll, having 

 once lost the leaflet character, would never regain it. 

 Of course we recognize reversions and similar phenomena 

 but regard them as restricted manifestations of heredity, 

 whose influence does not extend over any very great 

 period of time. For example, if some Paleozoic char- 

 acter should suddenly appear in a living species, we 

 should not attribute it to the influence of some long 

 dormant force of heredity but should regard it as a freak, 

 in no way due to the fact that some remote Paleozoic 

 ancestor may have had a similar feature. If we are 

 right in this opinion the living cycads could not have 

 been derived from the Bennettitales, like Cycadeoidea, 

 because the female sporophyll in these Mesozoic forms 

 had already lost more of the leaf character than have the 

 cycads of today. 



The reduction in the number of microsporangia on a 

 sporophyll and the reduction in the number of spores in 

 a sporangium furnish good illustrations of evolutionary 

 series. 



Embryogeny affords one of the strongest illustrations 

 of the drift of evolution within a group. Cycas and 

 Encephalartos both have a complete or nearly complete 

 segmentation of the egg during the early embryogeny, 

 but this does not mean that either inherited it from the 

 other; in fact, it seems probable that there was no such 

 inheritance in this case, but that both are still retaining 

 a type of embryogeny which characterized the ferns. 

 Other cycads, whose ancestors doubtless had this type 

 of embryogeny, have diverged from it to a greater or 

 less extent. Here again the superficial investigator is 



