THE ORIGIN OF FOREST-GROUPINGS. 237 



later Miocene and earlier Pliocene. There are the same species, 

 the same mixture of laurels, hollies, olives, to which are added 

 recent Japanese or Caucasian forms of nuts, maples, elms, and 

 toward the high summits pines and firs very like those of the 

 higher mountains of Teneriffe, Morocco, and Asia Minor. Collec- 

 tions made by M. Marion in the marly sediments of Durfort, 

 where bones of the southern elephant were also found, and in the 

 tufas of Valentine, near Marseilles, show that a number of plants 

 now found only further south, still in the second half of the 

 Pliocene inhabited the hills and shore-lines of southern France. 

 These plants were then, therefore, at home further north than 

 they are now found ; and their occurrence in the Pliocene points 

 to a later flowing back of species, amounting to a definite retreat 

 in the Quaternary age. 



This retreat, perfectly logical and almost regular in its opera- 

 tion, is connected with changes of climate, which were themselves 

 in relation with a progressive depression of the temperature of 

 the earth. It is connected also, in a parallel order of phenomena, 

 with the exhaustion of some races, and with the development, 

 by a concomitant origination, of other young and new races, 

 favored by the same circumstances that caused the elimination 

 of the races that gave way to them. Considering all the elements 

 of the question, we find that it is by the extension, at a given 

 moment, of vegetable races previously localized and realizing a 

 certain amount of variation, that species are constituted at the 

 start. Once characterized — that is, after the acquisition of a total 

 of characteristics, at first fleeting, then hereditarily fixed — the spe- 

 cies is permanent nevertheless only in a relative fashion so long 

 as there exist in it parts susceptible of differentiation anew. The 

 amplitude of the limits between which it may range through the 

 course of time depends on the proportion of the elements that re- 

 main variable to those that will not change. The morphological 

 oscillations of which it offers an example are thus determined 

 by its own tendencies to submit more or less readily to excita- 

 tions from without. Hence there are evident inequalities in the 

 specific type, sometimes running to obscure shades, sometimes 

 clearly cut ; the last especially after the exclusion of intermediate 

 forms. 



There exist, in fact, fleeting species, which can not be circum- 

 scribed by any precise limit ; and others, fixed in their minutest 

 traits, that are susceptible only of insignificant variations. The 

 forms of the latter category, such, for example, as the sequoias of 

 America and the cedars of the Atlas, persist in the places of 

 which they have once taken possession, where some of them have 

 been driven back and cantoned, so that quite contrary conditions 

 or the intrusion of more vigorous forms have not sufficed wholly 



