PROBLEM OP EVOLUTION — CAULLERY. 335 



tion of new forms. Hybridization among species, when it yields 

 fertile offspring, may, according to him, give rise, all at once, to a 

 whole series of new forms, whose mutual relations and differential 

 characteristics correspond exactly to what the natural species show. 



However subversive and delusive ideas of this kind, positive or 

 negative, appear to generations saturated with Lamarckism and 

 Darwinism, we must not lose sight of the fact that they were formu- 

 lated by eminent biologists, and that they are the result of long and 

 minute experimental researches and that many of the facts on which 

 they rest msiy be considered as firmly established. 



But without thinking of rebelling against the facts resulting from 

 genetic studies, we may question whether they have so general a sig- 

 nificance. I have already more than once pointed out that the 

 present aspect of organic heredity does not oblige us to conclude that 

 it has always been the same. We may ask ourselves w^hether condi- 

 tions, which have not yet been realized in experiment, do not either 

 modify directly the germinal substance itself, or the correlation 

 existing between the parts of the soma, and indirectly through them 

 the germinal substance. The facts which the study of internal secre- 

 tions are just beginning to reveal, perhaps indicate a possibility of 

 this kind. Even if we admit that evolution proceeds only discon- 

 tinuously bj^^ mutations, we still have to discover the mechanism of 

 the production of these mutations. In short, we may believe that, 

 with heredity and variations acting as recent researches have shown 

 them to act, there are nevertheless conditions that are still unknown 

 and that they have been realized for each series of organisms only at 

 certain periods, as seems to be suggested by paleontology, and in 

 which the constitution and properties of hereditary substances are 

 changeable. Of course these are purely hypothetical conjectures, but 

 such conjectures must be made if we wish to reconcile two categories 

 of already acquired data which we are obliged to recognize as facts. 

 On the one hand we have the results of modern genetics which of 

 themselves lead to conceptions of fixity, and on the other hand, the 

 mass of morphological data which, considered from a rational point 

 of view, seem to me to possess the value of stubborn facts in support 

 of the transformist conception; I will even go so far as to say in 

 support of a transformism more or less Lamarckian. 



It seemed to me necessary to devote the first meeting of the course 

 to this general analysis of the conditions under which the problem 

 of transformism now presents itself. I believe that this analysis is 

 the justification of the course itself. It shows the advantage of con- 

 fronting in a series of lectures the old classic data with the modem 

 tendencies, all of which have to be brought into agreement. The 

 crisis of transformism which Le Dantec announced some eight years 

 ago is very much more acute and more in evidence now than it was 

 then. 



