GYMNOSPERMtE. 35 



trusted, rather by the shortening, almost to abortion, of the upper and under leaves, than 

 through their being narrowed at the base and twisted, as at present, towards the sides of 

 the branchlets. 



The Arctic Tertiaries have yielded no foliage of the S. gigantea type, except that 

 referred erroneously to 8. CoutfsicB, which I distinguish as S. Whymperi} The S. sem- 

 pervirens type, on the contrary, was exceedingly abundant; S. Langsdorjii being, in fact, the 

 prevailing fossil in the Greenland Tertiaries, for scarcely any stones with leaf impressions 

 are without some traces of it. The branchlets seem to have been shed simple or singly, 

 rarely forking, and would thus perhaps appear to have had but a short season of growth, 

 an adaptation possibly to Arctic conditions. 8. Langsdorjii again appears in the Miocene 

 of the Baltic, in ]\Iull, in the Mayencian or Lower-Miocene stage in Switzerland and 

 Germany, and in Italy in the latest Miocene period. Its course southward from the 

 Arctic Eocenes to Italy, as the temperatures gradually decreased in the Miocenes, is 

 therefore indicated, like that of so many other Coniferse. The branchlets from the more 

 southern deposits, it is curious to notice, are sometimes compound. Another Spitzbergen 

 species is distinguished by smaller, more tender foliage, and narrower leaflets set at a more 

 acute angle with the stem. 



The Eocenes of Erance have yielded two well-marked species, one very distinctly, and 

 the other to a less degree dimorphic. Another extremely dimorphic form, 8. Hardtii} is 

 abundant at Steiermarkt and Haring in Austria. 8. Whymperi seems certainly to have 

 been dimorphic, and to have descended south towards the Miocene. 8. brevifolia is not 

 dimorphic in Greenland, but becomes so in the American Eocene, and even in the most 

 persistent of the distichous types, 8. Langsdorfd the fruit-bearing branches are imbricated 

 to a greater extent than is usual in 8. sempervirens at the present day, though in trees culti- 

 vated in warmer climates, as Madeira, the imbricated foliage is relatively much developed. 

 The purely imbricated type is rare in the Tertiaries. After mature consideration the 

 so-called 8. 8ternhergi of Bournemouth is placed in Araucaria, and I have failed to dis- 

 cover any reasons why the greater part of the similar foliage so widely spread in mid- 

 European Eocenes should not also be transferred back to the same genus. In some 

 cases it might belong to Sequoia, but no Sequoia cones have yet been found with it, and 

 until some better evidence is forthcoming it is preferable to consider all the Eocene foliage 

 of this kind as belonging to one genus. In the Iceland Tertiaries, however, there is an un- 

 usually large-foliaged 8equoia, confounded with 8. Sternbergi, though it is very distinct 

 in character from the typical form. This reappears in the Superga of Turin, being in both 

 cases associated Avith cones, and apparently also at Sinigaglia {Pinites crgptomerioides, 

 Mass.), and in the Great Lignitic of America {8. acuminata, Lesq.). 



The striking difference in the form of the foliage belonging respectively to the two 

 recent species is seen to have been far less pronounced in the past ; even now 8. sem- 



1 See S. Cm It sice, p. 39. 



" Long known as Chumoecyparites Ilurcltii. 



