Evolution of Plants 361 



examples, or their still more primitive ancestors. But, while 

 the stem structure, the flowers, and the fruit of the Ginkgoales 

 and Taxales show many marked points of affinity, the con- 

 densed leafage of the latter, and the tertiary spiral thickening 

 that is so marked a feature of the fibers in the yew alliance, 

 indicate that during the early permian, in all probability, a 

 more or less marked separation of genera from the Ginkgoales 

 took place, and from such derivative genera in turn evolved 

 the taxaceous organisms of today. 



During the process the gradual absoiption of the circlet 

 of cilia round the spermatozoids, that is still present in Ginkgo, 

 must have taken place, and reduction of the resulting speim- 

 nuclei to rounded masses that pass along the pollen-tube and 

 toward the egg-cells by chemotactic action and osmosis was 

 simultaneously effected. 



The Protopinales or primitive Coniferales form a division 

 that can recognizably be traced back as such to Jurassic times. 

 It seems highly probable also that their ancestral connections 

 are to be met with in such cordaital genera of the coal period 

 as Cordaianthus, in which staminate and pistillate flowers form 

 close spicate masses, that are largely made up of imbricated 

 sporophyll leaves. 



The most ancient group of them now living, and the group 

 also which appears to have made closest contact with cordaital 

 predecessors, is the Araucariese. In truth so strikingly does 

 the wood in the often large stems of carboniferous cordaitals 

 resemble that of existing Araucarias that it was called Arau- 

 carioxylon by the eailier palseophytplogists of fifty years ago. 

 The leaves also of Agathis, and of some species of Araucaria 

 now living, conform closely to cordaital affinities. 



But in discussion of the phylogeny of Coniferales it may 

 be well to refer here to the extreme diversity of foliage in the 

 one genus Araucaria. Thus the leaves on vigorous shoots 

 of A. imhricata recall greatly those on seedling shoots of Agathis 

 (Damrnara), while histologically they show not a few points 

 in common. Those of A. hrazilinesis and A. Bidwillii accord 

 more nearly with the genus Cunninghamia; A. Cookii resembles 

 Abies or Picea types; while A. excelsa in general morphology 

 and histology approaches to cupressinoid characters. In other 

 words this ancient genus, through long ages of adaptive variation 

 in part in the northern but mainly in the southerrr hemisphere, 

 has undergone modification, that greatly duplicates the varia- 

 tions shown by leading divisions of the entire group Corriferales. 



The genus Voltzia that extends upward from permian to 

 the Keuper beds, and Walchia that shows close affinities with 



12* 



