VO l889 n 'J PROCEEDINGS OF THE NATIONAL MUSEUM. G05 



This description coincides, it will be observed, with the descriptions 

 of species of Dadoxylon, particularly with what Williamson has de- 

 scribed as a typical Dadoxylon.* Indeed, there can be no doubt that 

 some, at least, of the Paleozoic forms referred to Dadoxylon really repre- 

 sent the wood of Cordaites. 



But the classic memoir of Renault, " Structure Comparee a quel- 

 ques tiges de la Flore Carbonifere," which appeared in 1870, must 

 remaiu the most valuable contribution to our knowledge of the genus 

 Cordaites. From a study of the very perfectly preserved specimens 

 collected in the vicinity of Atun and St. Etienue, he was able to make 

 out the cellular structure of the immense pitb, of the medullary rays, 

 of the wood cells showing the hexagonal arasolatiou, and also of the 

 very thick bark. Likewise the structure of the leaves, the develop- 

 ment of the male flower, manner of fertilization, and growth of the 

 ovule were clearly described, so that we actually have a more complete 

 knowledge of this long-since extiuct genus than of many living genera. 

 Renault proposes to place it as a distinct family between the Cycads 

 and Conifers, being related, as Schenck has pointed out, by the male 

 flowers to the Salisburiea^, by the female flowers to the Cycads, while 

 in the structure of the wood it can hardly be distinguished »from the 

 Conifers. 



It being now acknowledged that there is a genus showing Araucaria- 

 like structure, yet unquestionably distinct from it, it becomes neces- 

 sary to inquire if there are characters which can be used to distinguish 

 them when there is not a sufficient amount of material to settle this 

 macroscopically. According to Grand'Eury and Renault, and also by 

 Morgenroth, Vater, and Felix, who have taken up the subject within a 

 few years, there undoubtedly are characters that may be relied upon 

 to separate Cordaites from the other Paleozoic woods. The possession 

 of an Artisia pith is a clear indication that the example belongs to 

 Cordaites, since, as stated above, Grand'Eury has always found the 

 Artisia pith associated with Cordaites when the specimen has been 

 entirely preserved. In all the species that have been described by 

 Renault, Felix and others as typical, the punctations have entirely 

 covered the radial walls of the tracheids. There are several other 

 minor characters that will be enumerated when we come to the detailed 

 description of the species. 



Having settled that Cordaites must be separated from the genus 

 Araucarioxylon of Kraus, the further question arises as to what is to 

 be done with the remaining forms. According to the best recent 

 authorities, as Sheuck, Felix, Lesquereux, Morgenroth, and Fontaine, 

 the true Conifers of the Araucariau type do not extend into the Paleo- 

 zoic. The AraucariaB first had their origin in the Jurassic, or in any 

 case in the Mesozoic, and here the trunks are for the first time found 

 associated with undoubted leaves and cones of the Araucariau type. 



'Organization Foss. PI. Coal-meas., vm, p. 222. 



