EemarMon the Ancient Flora of the Earth. 113 



been done among the incomparably more numerous tribes of 

 the animal kingdom belonging to a former age. 



In consequence of the nature of the subject, investigations of 

 this kind necessarily consist of two parts essentially different from 

 each other, and the comprehensive knowledge they require is sel- 

 dom found united in the same individual. For it is required not 

 only to reconstruct the whole of an organic body from the 

 imperfect remains that are preserved, and hence to draw a 

 conclusion as to the family or species in which it may be classed,— 

 but it is equally necessary to determine the nature and the age 

 of the rock formations in which these organic remains are found. 

 Further, the first part of the investigation referred to, which 

 belongs entirely to the province of Natural History, is incom- 

 parably more difficult to be developed from the remains of vege- 

 table bodies than from those of the animal kingdom ; for the 

 essential characteristics of the latter are more numerous, and 

 much less liable to complete destruction. The zeal with which 

 our cotemporaries have pursued this object, has greatly contri- 

 buted to remove many of the difficulties ; and, independently of 

 the first attempts, particularly those of Schlotheim and Count 

 Sternberg, no one has devoted himself to these pursuits with 

 greater sagacity and success than Brongniart. The natural 

 history of fossil plants, in almost all its parts, has been com- 

 pletely changed by this eminent naturalist, and in a wonderfully 

 short time. Though we are anxious to make our sincere ac- 

 knowledgments of the great value of the descriptive part of his 

 Natural History of the Vegetable Kingdom, we are, notwithstand- 

 ing, obliged to put forth some objections to the purely geological 

 part of M. Brongniart's work, which occurred to us in the course 

 of reading his introductory treatise.* 



After the character of the Floras into which the vegetation 

 ofa former age may be divided, M. Brongniart distinguishes 

 four different periods, and he determines them geognostically 

 in a similar though much more accurate manner, than had 

 formerly been attempted by Count Sternberg. (Fascic. iv. 

 p. 32.) According to him, the first period comprehends the 



• In the Edinburgh New Philosophical Journal, vol. vi. there is a transla- 

 tion of Brongniart's Memoir " on the Vegetation of the Earth at different 

 periods." 



OCTOBER — DECEMBER 1829- H 



