186 CONIFERS. 



kept in mind in the identification of fossil specimens. It may be 

 useful to draw attention to similar difficulties in the case of 

 Comfercs, which have not always been observed by palaeobotanical 

 writers. 



If we examine the external characters of older branches of recent 

 conifers from which the leaves have been detached, it will be found 

 impossible to institute on such a basis any useful classification. It 

 happens, not infrequently, that the leaves and cortical tissues 

 become readily detached from the surface of the wood, leaving 

 a smooth axis in place of the corticated branch or stem. A good 

 example of this is afforded by such specimens as those represented 

 in PI. XVII. Figs. 4-6. Occasionally we have to deal with pith 

 casts having the surface covered with lozenge-shaped prominences, 

 simulating elongated leaf bases. A good example of such a 

 medullary cast is afforded by Weiss' genus Tylodendron, of which 

 the true nature was pointed out by Potonie * in 1887. Again, in 

 some specimens of the Triassic Volfaia i we have smaller pith casts 

 of similar form. In his Introduction to the Study of Palceontological 

 jBotamj, Balfour 3 calls attention to the unnecessary multiplication of 

 fossil species, and illustrates the need for careful observation of the 

 characters of recent stems, by reference to the striking differences 

 presented by a branch of Araucaria imbricata, Pav., when the bark 

 is viewed intact, and after it has been more or less completely 

 stripped off the surface of the wood. In Araucaria Cunninahamt, 

 Ait., we find equally striking contrasts between the younger 

 branches, with their stiff falcate leaves, the slightly older stems, on 

 which only the rhomboidal leaf bases are left, the smooth surface of 

 the wood, from which overlying tissues are readily detached, and 

 finally, the surface features presented by a pith cast. 



In the long needles of Pinus and the broad flabellately veined 

 leaves of Ginkgo, we have sufficiently well-marked characters to 

 enable us in most cases to arrive at a generic determination. In 

 many instances, however, it is a hopeless task to attempt to found 

 any accurate determination on leaves alone. Among recent genera 

 we have a deciduous habit in such plants as Larix, Ginkgo, Taxodium 



1 Potonie (2). 



2 Seward (4). 



3 Balfour, p. 4 



