606 REVISION OF ARAUCARIOXYLON KNOVVLTON. 



The so-called Araucarites gracilis, described by Dawson from the Car- 

 boniferous and Permian of Canada, is without much doubt a species of 

 Walchia, a well known Permian genus. 



As the Mesozoic and Tertiary forms are manifestly the only ones 

 that can be justly referred to Araucaria, Felix has proposed* to con- 

 fine the use of Araucarioxylon of Kraus to woods from these forma- 

 tions. This seems on the whole a desirable chauge and quite in har- 

 mony with modern views. 



There now remain the Paleozoic forms not included in the genus 

 Cordaites. As they can not of course be called either Cordaites or 

 Araucarioxylon, Felix has proposed! to restore Endlicher's name — 

 Dadoxylon — for them. The term Dadoxylon. meaning simply ancient 

 wood, is a very good one for these woods, inasmuch as it is non-com- 

 mittal, and the former may be changed from it at any time provided 

 their true relationship may be made out. They are probably to be 

 regarded as the ancestors of the true Araucarire, although the dis- 

 covery of more perfect material may possibly show them all to be re- 

 lated to or included in the Cordaiteae. 



CORDAITES t Unger. 

 1. Cordaites Ouangondianus (Dn.) Gopp. 



Goppert in Nachtriige z. Kenntniss d. conifereuholzer d. Paleoz. Form., 1888, p. 9. 

 Dadoxylon Ouangondianum Dawsou, Can. Nat., Vol, vi, 1861, p. 165, figs, (in 



text) 1-4. 

 Araucarioxylon Ouangondianum (Daws.) Kidston. Cat. Brit. Mus., 1886, p. 237. 



Branching trunks with distinct zones of growth and a pith of Artisia 

 (Sternbergia) type; wood cells very large, with three to five rows of 

 contiguous alternate hexagonal areoles with oval pores; medullary 

 rays with one to three series of cells aud as many as fourteen rows 

 of cells superimposed on each other (Dawson.) 



Middle Erian (Devonian) formation, of Canada. 



2. Cordaites Halli Dn. Sp. 



Dadoxylon Halli Dawson. Quat. Jour. Geol. Soc. Lond., Vol. xvm, 1862, p. 

 306, PI. xiii, fig. 11 ; Foss. PI. Devonian and Silurian Formations of Can- 

 ada, 1871, p. 14, PI. I, figs. 5,6. 



Wood cells very large, with five rows of contiguous alternate hexa- 

 gonal areoles ; medullary rays very frequent, and with as many as 

 thirty rows of cells superimposed. (Dawson.) 



Hamilton group, Hemlock Creek, Ontario County, New York. 



*Abhandl. d. k. geol., Landes-Anstalt, Bd. VII, 1886, p. 209. (Page 57 of reprint.) 



t Op. tit., p. 209. 



t Felix and others write Cordaioxylon, but in view of what has been brought out 

 in the preceding pages regarding the completeness of our knowledge of the life his- 

 tory of Cordaites, it does not seem expedient to use different generic names to desig- 

 nate simply different parts of the same plant. Goppert in his last work also 

 adopted this view and wrote Cordaites for the species founded upon internal struc- 

 ture, as well as for those founded upon other parts of the plants. 



