12 NATHORST. FOSSIL PLANTS OF FRANZ JOSEF LAND. [norw. POL. EXP. 



Nevertheless it must be admitted there is a very great resemblance be- 

 tween Ginkgo polaris and the two species mentioned, so that there is the 

 temptation of classing some fragments with the one species, some with the other. 

 As, however, the most perfect specimen of Ginkgo polaris does not agree 

 with either, I have deemed it more correct to consider it as a species of its 

 own, more particularly as no one of the many specimens examined by me, or 

 by Newton or Teall, presents so long and strong a petiole as those possessed 

 by the species from East Siberia. 



To Ginkgo polaris we must therefore refer the specimens represented 

 by Newton and Teall in pi. 38, figs 4 and 5, which they call Ginkgo sibi- 

 rica. If they had belonged to the latter species, at any rate the specimen 

 fig. 5, should still have shown the long petiole of the leaf. The specimen 

 figured in their second paper, (plate 29, fig. 3) and which they with hesitation 

 refer to Ginkgo polaris, is also characterised by a very short petiole. 



Ginkgo polaris Nathorst var. pygmcea n. var. 



PI. I. figs. 20, 21 and (magnified) 50, 52. 



It is probable that these specimens are only a variety of the preceding, 

 since transition forms do not seem wanting (fig. 18). The specimen fig. 20 

 is the very smallest of all the Ginkgo leaves hitherto described, as it is even 

 smaller than Heer's Ginkgo pusilla from the Jurassic strata of East Siberia, 

 which itself is very closely allied to Ginkgo flabellata. The specimen in 

 question has its leaf divided into four lobes, while the fragmentary specimen 

 fig. 21 is divided into six, but with signs of a commencing division of the 

 two innermost lobes, which are consequently broader than the rest. The veins 

 in the specimen fig. 20 are five in each lobe, in the specimen fig. 21 similar 

 in number in the narrower lobes, but nine to ten in those that are broader. 



Ginkgo sp. 



PI. I. figs. 22—24. 



In my opinion these specimens belong to a separate species, as they differ 



from Ginkgo polaris by having a larger, less deeply divided lamina, more 



distmct venation and often present a peculiar structure (fig. 23a enlarged) 



which seems to correspond with transverse cracks in the carbonised leaf sub- 



