species in the case of certain Fishes. 439 



Whether it helongs to a single primitive tree or to one of the 

 descendants of the latter in the forest of beings , the hud, species 

 or variety, which terminates a branch must always possess the 

 poicer of yielding more or less to the exigencies of a variable 

 medium, and of being able to produce thus new modifications of 

 greater or less importance, themselves endowed, in their turn, 

 with greater or less variability and vitality. 



Most authors who desire to give an absolute definition of 

 the species, generally invoke, as evidence of stability, the 

 difficulty of intercrossings between distinct species, and the 

 comparative sterility of the hybrids thereby produced, as also 

 the facility with which, on the other hand, the races derived 

 under our eyes from a single stock multiply together. These 

 difficulties, however, which are often exaggerated, frequently 

 seem to result from the desire to unite, for certain advantages, 

 organisms endowed with useful qualities of too opposite a nature. 

 In the two cases we are at a very different distance from the 

 parent form ; it is necessary, as Besnard has already indicated, 

 to be able to make one's choice, or to return further back in 

 the ramifications of the genealogical tree. It is probably for 

 an analogous reason that it is usually the. lower types that pre- 

 sent most possible modifications or combinations. A longer 

 duration of the influences, by more profoundly altering the 

 oi-ganisms, evidently diminishes the sentimental attraction, if 

 we may so call it, that a similarity of appearance must neces- 

 sarily favour, and at the same time renders a perfect combi- 

 nation of the organism of the two individuals selected less 

 easy to be effected in a manner sufficiently complete to become 

 productive. 



It is impossible not to see here the existence of two general 

 laws opposed to one another and constantly struggling one 

 against the other, and which, according as they are called by 

 circumstances to predominate over one another, maintain the 

 species within relatively immutable limits, or, on the contrary, 



of a single step. In other words, I cannot decide whether the genealo- 

 gical tree of living creatures was planted with all its smallest branches, as 

 Agassiz thought, or whether a primordial cell, in place of a seed, origi- 

 nally gave birth to a genealogical tree which, at first an aquatic plant, 

 gradually extended its branches upon tlie solid ground, and, growing 

 larger and larger, put forth all the branches which uow-a-days constitute 

 the totality of the known and unknown organisms on the face of the 

 globe, according to the views of some disciples of Darwin, RoUe, Hackel, 

 and others. It matters little to me, in fact, in the ascertainment of the 

 variability of an existing species, whether I assume the existence of one 

 or several seeds, whether 1 see u single tree constantly increasing in si/e, 

 or perhaps still believe in the existence of a whole forest of genealogical 

 trees, sprung fi-om the seeds of a single plant, but from germs which have 

 fallen successively under dift'erent conditions. 



30* 



