r.< I :•..'.. II. f.. CLAPSIl rCATlON (iK FOSSIL I'l^ANTS. 371 



iind the (nic most often exj)r»'sse(l, is that this cniirsf imolves 

 the creation of lari^^e niunhers of transient speeitic names, 

 due onh' to the mistakes' of the respective authors. Havini; 

 r^'gard to thr limited nature of the i)alaeobotanical material, 

 it nf'»'(l hardly be feared that the mere nunilM-r of names will 

 cause serious troubles. And there is this very important 

 <litterence between the two eases: that a laro^e number of 

 name-, which are bound each to a well-defined form, when 

 finally acknowledged as mere synon^'ms of a species brini; 

 murh less confusion than a much smaller number of others 

 which are used in a different sense by different authors and 

 have Ijeen applied each to different species. The splitting up 

 of a vaguely defined species will always be found much more 

 <lifficu]t than the joining of tw(t narrowly defined forms to 

 one species and the latter process only represents a natural 

 stage in the progress of the classification of fossil plant-remains. 



The phytogeographical side of the question has been for 

 the present sufficiently considered in my paper on the Gra- 

 ham Land Flora, but it remains to add a few words in regard 

 to its bearings on the stratigraphy. The importance of not in <; 

 tliH exact geological horizon on which a certain form occurs 

 lia< been emphasized by Prof. Xathorst — who gives, more- 

 over, in his paper on the naming of dicotyledonous leaves* an 

 exposition of many of the most important precautions to be 

 considered in the classification of fossil plants. Kven a small 

 difference between two forms may be of importance if each 

 form is found to be restricted to a certain horizon. A closer 

 attenti<m to this question than is generally bestowed b\' 

 palaeol)otanists may often be found useful in the delimitation 

 of species, and, rice rcr.sa, may increase the geological impor- 

 tance of fossil plants. For this purpose, however, a narrow 

 and careful deliniital ion of species is of itaraniount im])ortance. 



It is not intended to advocate, with the above remarks, an 



' Xathorst. A. <J rdier die Hentnnung f<issil«T l»iki>tylonhlatter. Hota- 

 nisches Centraildatt. Vol. 2^). iMSfi. 

 \t^ — i:i022U. G. F. /•'. /!'/.y. 



