Species in the case of certain Fishes. 441 



Although called upon to vary, more or less, and with time, 

 in certain countries, when in spreading it has met with new 

 exigencies, the species may nevertheless remain relatively 

 fixed, or vary comparatively little in the same locality or in 

 analogous media, so long as the conditions are not sufficiently 

 modified. This is what led the illustrious Cuvier to say, and 

 up to a certain point with justice, " Experience seems to 

 show, on the contrary, that, in the actual state of the globe, 

 varieties are confined within very narrow limits ; and so far 

 as we can go back into antiquity, we see that these limits 

 Avere the same as at the present day." 



I have already several times recognized and indicated, in 

 certain widely distributed species belonging to various classes 

 of our Vertebrata, nascent divergences in some part or other 

 of the animal. These variations, more and more strongly 

 marked until tliey reach adaptation through the persistence of 

 the influences and heredity, constitute what I call tendencies, or 

 the origin of new bifurcations upon a genealogical branch. 

 Often perceptible in certain individuals in a very limited field 

 of observation, they increase more and more in other coun- 

 tries with the augmentation of the first small dissimilarities 

 of condition, and thus advance towards temporary maxima^ 

 which in various places have received different specific names. 



The origin of these divergences may be, according to cir- 

 cumstances, attributed to a persistence of the characters of 

 youth, or to the predominance of the distinctive features of 

 one or the other sex, or to the reproduction by heredity of a 

 quasi-accidental anomaly, or, again, in consequence of the 

 struggle for existence, to new exigencies of the conditions of 

 life. I have particularly indicated, in the number of the 

 ' Archives de la Biblioth^que Universelle ' for September 

 1876, the coexistence, in the waters of the Lake of Geneva, of 

 three very distinct tendencies in the forms ot the roach [Leu- 

 ciscus rutilus). Each of these three varieties (deep, elon- 

 gated, or thick) already indicates, with a primary modification 

 of the general form, more or less strongly marked correlative 

 deviations in several of its characters. 



Without going beyond the restricted bounds of our own 

 ichthyological fauna, I might cite several other cases of varie- 

 ties of one and the same species living thus almost side by 

 side, although kept distinct by exigencies of medium, which 

 are often badly interpreted. It may suffice for me, in this 

 connexion, to refer to the example of our freshwater trout, 

 which, according as it is more or less confined to small streams 

 or to the deeper waters of our lakes, presents a facies so diffe- 



