496 



SCIENCE. 



[Vol. V., No. 124. 



question, with the abundant and so long enig- 

 matic Cordaites. This ancient plant was for- 

 merly regarded as the forerunner of the family 

 of cycads ; but now, in the light of these dis- 

 coveries, it is almost universally regarded as 

 coniferous. It was one of the earliest types of 

 land vegetation to appear on the globe, running 

 far back into Devonian, and even into Silurian 

 time. 



The figures of the accompanying plate, 



remarkable manner the almond-shaped nuts 

 borne by the present maiden-hair tree. 



Though these carboniferous plants were at 

 first commonly regarded as cycadaceous, still 

 the long, ribbon-like leaves of certain cordaitean 

 forms (Poa-Cordaites of Grand' Eury) led some 

 eminent authors, including the late Professor 

 Goppert, to consider them monocotyledonous, 

 and the precursors of our lilies, reeds, grasses, 

 and also of the palms. But even these mis- 



PHYLOGENY OF THE GENUS GINKGO. 



Cordaites lingulatus, Grand'Eury : carboniferous, Central France. 2. Psygmophyllum (Noeggerathia) flabellatum (Lind. & 

 Hutt.), Schirap. : carboniferous, England. 3. Ginkgophyllum Grasseti, Saporta: Permian, Herault. 4. Baiera (Jeanpaulia) Miin- 

 steriana (Presl), Heer : rhaetic, Bayreuth. 5. Ginkgo Sibirica, Heer : oolite, Siberia (restored by Heer). 6. Ginkgo digitata 

 (Brongn.) , Heer : oolite, Cape Boheman, Spitzbergen. 7. Ginkgo Laramiensis, n. sp. : Laramie, Point of Rocks.Wyoming Territory. 

 8. Ginkgo (Salisburia) adiantoides, Unger : Fort Union beds, lower Yellowstone. 9. Ginkgo (Salisburia) adiantoides, Unger : 

 miocene, Greenland. 10. Ginkgo biloba, L. : living, Washington, D.C. All the figures are reduced one-half. 



kindly drawn for me b}' Ensign Everett Hay- 

 den, U. S. navy, have been selected with a 

 view to illustrating the phylogeny of the 

 genus Ginkgo: and the} r are numbered, and 

 as nearly as practicable arranged upon the 

 plate, in the order of supposed development, 

 from the true Cordaites to the living Ginkgo 

 biloba ; this being also, as will be observed, 

 substantially the chronological order of their 

 appearance. 



The broad leaves of some species of Cordai- 

 tes, though more or less elongated or elliptical 

 in shape, possess a nervation strikingly similar 

 to that of the later ginkgo-likc forms ; while 

 the familiar fruits so abundant in the coal- 

 measures, and which are now known to be 

 those of Cordaites, resemble in an equally 



takes have not been without their uses. It is 

 the peculiarit} T of science that in its very errors 

 knowledge is extended. The theory that Cor- 

 daites was c}*cadaceous was not wholly false ; 

 the suggestion that it might be monocotj'leclo- 

 nous contained a ' soul of truth ; ' and the pres- 

 ent opinion that it was coniferous is, I venture 

 to assert, not wholly true. The truth lies in 

 the midst of all these opinions. It seems to be 

 this : there were no true paleozoic Cycadaceae, 

 monocotyledons, nor Coniferae; but Cordaites 

 was the prototype of them all. It was in the 

 Trias, whose flora is unfortunately the least 

 known of all the formations in past time, that 

 all these definite types of vegetation were dif- 

 ferentiated from this comprehensive type, — 

 the Cycadaceae through their Macropter3'giums 



