342 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1918. 



Nilssoniales or Diooniales, effectually obscures the fundamental plan 

 of evolution of the phylum, if its history has been at all like that just 

 outlined. The name Williamsoniales is unfortunate, and probably it 

 would be better to use Hemicycadales (Wieland). 



As sketched in the following paragraphs the Cycadophyta are 

 somewhat arbitrarily segregated into three subordinate groups: The 

 aberrant Cycadeoidales, which are not as far removed from the second 

 group, the Mesozoic and existing Cycadales, as formerly was thought 

 to be the case. The third group, for which the inappropriate name 

 Williamsoniales is retained, includes the balance of the Cycadophytes. 

 It is not a compact group, and the progress of discovery will doubt- 

 less show it to be unnatural. In its beginnings it approximates the 

 Pteridospermophyta and also shows points of possible contact with 

 the Gnetales and Ginkgoales. In its more evolved members (e. g., 

 Williamsonia) it approximates the Cycadeoids to such an extent that 

 Wieland and other authorities group the former and the latter to- 

 gether. The two points of view are entirely in agreement regarding 

 the facts and differ merely in the placing of the boundary between 

 the two groups, a not very serious difference in an evolving series of 

 forms. 



Although in Cycadeoidea Gibsoniana, the species in which the floral 

 organization was first made known, only ovulate organs were dis- 

 covered, the much fuller American material leads to the inference 

 that the flowers were normally bisporangiate. 



The Cycadeoidales appear to represent an evolution from the older 

 Williamsoniales stock, and according to our present knowledge, they 

 are confined to the Jurassic and Lower Cretaceous. Their prevail- 

 ingly tuberous trunks are individually abundant in a silicified condi- 

 tion in certain sandy strata of those ages, particularly in the Portland 

 Dirt Bed of the Isle of Wight, in the Patuxent formation of Mary- 

 land, and the Morrison beds of the Freezeout Hills in Wyoming. By 

 far the largest number have been found in the similar strata of the 

 Lakota formation, which outcrops in the Black Hills rim. Similar 

 remains in less abundance have been found in France, Italy, Galicia, 

 and elsewhere. Their wonderful preservation and exceptional mor- 

 phological features have resulted in unusually painstaking researches 

 which have shown that the group as a whole was one of rather limited 

 variation and stereotyped organization, relatively no more abundant 

 in Jurassic and Lower Cretaceous times than are the Cycads in exist- 

 ing floras. 



The salient features, briefly enumerated, are a tuberous or short 

 columnar stem, sometimes an aggregate of tuberous stems, as in some 

 species of Encephalartos. The stem was encased in a heavy armor 

 of persistent leaf bases, which largely obviated the necessity for the 

 development of a thick zone of mechanical woody tissue in the stem. 



