296 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1918. 



reversible. The earlier forms of the successive plant groups, in 

 accordance with the a priori conclusions of evolutionary philosophy, 

 are in general synthetic or generalized in structure, i. e., they com- 

 bine features of various categories that subsequently became charac- 

 teristic of diverging lines of evolution. Thus all of the earlier plants 

 had motile sperms; none had true vessels such as characterize the 

 wood of angiosperms ; cryptogamic, or centripetal wood formation 

 was an almost universal feature, and many other similar characters 

 could be mentioned. The Sphenophyllales were synthetic in that 

 they combined certain characters of their fern ancestry with features 

 that subsequently became stereotyped in the Lepidophytes and Cala- 

 mites; they are thus considered as representing an approach to the 

 common ancestor of the balance of the Lepidophytes and of the 

 Arthrophytes. The earliest ferns show combinations of features 

 that subsequently became the property of different fern families; 

 while the Pteridosperms combined the characters of ferns and 

 cycadophytes. 



Another principle illustrated by paleobotany is that known as 

 recapitulation, or the fact that, theoretically at least, the ontogeny 

 or development of the individual is an epitomy or abridgment of the 

 phylogeny or development of the race. This principle is, however, 

 difficult of application since its underlying factors are unknown and 

 the ontogeny may show acceleration, modification, or obsolescence 

 that renders the application of recapitulation questionable unless 

 corroborated by other lines of evidence. 



A familiar instance of the retention by the seedling of ancestral 

 adult characters is furnished by the conifers, most of which have 

 needle leaves on the shoots of the juvenile forms, while with age 

 these needle leaves, of the long shoots are replaced by scales as in 

 an adult pine, or become reduced to concrescent leaves with a cyclic 

 instead of a spiral arrangement as in the juniper, or become reduced 

 to scales while the shoot (phylloclad) functions as a leaf as in Phyl- 

 locladus. From these changes in foliage in the earlier part of the 

 life history of the various conifers it is inferred that the scale leaves 

 of Pinus, the double leaf of Sciadopitys, the phylloclad of Phyl- 

 locladus, the concrescent cyclic leaves of the Cupressaceae and the 

 deciduous habit of the larch (Larix) and bald cypress (Taxodium) 

 are all derived characters. 



Another principle is derived from the study of persistence of 

 organisms or organs. Such conservative organisms as the Bacteria 

 or the Cyanophyceae have existed unchanged for millions of years. 

 Conservative organs are those like roots, whose functions and en- 

 vironment have remained uniform for ages. Thus all roots are 

 found to be much alike in their organization. Various enthusiasts 



