226 THE AMERICAN NATURALIST [Vol. XLIII 



retained more primitive characters in its sexual reproduc- 

 tion and vascular system; another line (Cordaitales), 

 while retaining the more primitive sexual reproduction, 

 developed a more advanced type of vascular system, 

 which has continued in Ginkgoales in one direction, and 

 in Coniferales in the other, associated in the latter with 

 a more advanced type of sexual reproduction. 



This gymnosperm situation may illustrate a fact that is 

 becoming more and more apparent. On the basis of the 

 older reproductive morphology the cycads are more 

 ancient than the conifers; on the basis of history the 

 reverse is true. The cycads are relatively modern, but 

 have persistently retained certain ancient features; yet 

 the logic of the older morphology would have insisted that 

 conifers are derived from cycads. It is a fact, therefore, 

 that primitive features are not necessarily a mark of age, 

 even among closely related groups. The testimony of 

 all features must be considered, and this checked up by 

 history, before any rational conclusion can be drawn. 



The most baffling tangle of relationships among gymno- 

 sperms at present is that presented by the tribes of Conif- 

 erales. The perplexity of the situation is due to the fact 

 that as yet morphology, vascular anatomy and history 

 are at variance. Of course, history must determine the 

 actual sequence, and then our contradictory morphology 

 and anatomy can be straightened out. For example, 

 morphologically in Taxinea? are advanced; in wood 

 structure they are also said to be advanced ; but they are 

 also reputed to retain the old mesarch structure, which 

 would indicate that they are primitive. When we know 

 what they are historically, we can determine whether this 

 anatomical feature is primitive because the group is 

 primitive, or because this character of the bundle has 

 lagged behind. Of course, these " mesarch' ' bundles may 

 not be mesarch in the old sense, and the centripetal wood 

 may be explained away ; if so, the group will be advanced 

 in all its characters and will not need the testimony of 

 history. 



