532 



NATURE 



[April 7, 1898 



however, is a matter of secondary importance ; Roux 

 insists specially on the use of experiment — accurate 

 painstaking experiment. — in biological investigation. He 

 further indicates that developing organisms afford the 

 most fruitful field for the experimental method, for there 

 one may most certainly hope to discover the formative 

 forces which by their interaction co-operate to produce 

 those formal changes which we have come to know by 

 the method of simple observation. It is on this subject 

 that Hertwig differs most widely with him. According 

 to the latter author, there is no place for the experimental 

 method in embryology. Experiment is nothing more 

 than the production of changes of state in existences. 

 In the inorganic world we have to deal with relatively 

 stable existences, and before we can make any assertions 

 of cause and effect about them we must bring about 

 a change of state in them. In the organic world, 

 however, the case is widely different. It is the character- 

 istic of living bodies that they are always undergoing 

 changes of state, and the changes are most characteristic 

 and most conspicuous during the period of embryonic 

 development. Thus nature does for man in the organic 

 what he himself has to effect in the inorganic world, and 

 it is only necessary for him to observe and record the 

 natural successive changes in order to be able to state a 

 series of relations of antecedent and consequent. Thus 

 Hertwig says — 



" Every antecedent state is the cause of that which 

 follows it ... a living frog's ovum is the antecedent 

 which of invariable necessity leads to the establish- 

 ment of a frog's gastrula as a consequent, if only 

 the conditions and circumstances necessary to further 

 development are fulfilled. For the words antecedent 

 and consequent one may equally well substitute 

 the words cause and effect. Hence embryological 

 research, which 'describes' the change of the frog's 

 ovum into the gastrula, asserts a causal relation, and in 

 so far as it does this for all the stages of the develop- 

 ment of the frog from the egg, it asserts the law of the 

 development of the frog. In this sense the research of 

 the last fifty years has brought to light the most im- 

 portant causal knowledge. Is not the recognition that 

 the ovum and the spermatozoon are simple elementary 

 organisms and that, as such, when the appropriate 

 conditions are fulfilled, they unite in themselves all the 

 causes (exception being made of caus<2 externce) which 

 are necessary to the production of a new being, and that 

 they in fact bring it into existence, is not this a causal 

 recognition ? " 



The above paragraph is quoted by Roux as illus- 

 trating very clearly the difference between his and 

 Hertwig's standpoints. Hertwig imagines that the ends 

 of science are'/ulfiUed by the enumeration and description 

 of different states, and holds that our task is finished 

 when we are able to assert that any one state invariably 

 proceeds from another state immediately preceding it. 

 Roux admits the necessity and value of this knowledge, 

 but declares that it is only a step towards a causal 

 explanation of the phenomena, and is far from satisfying 

 our desire for a full explanation. 



An illustration will serve to make the point clear. 

 Hertwig's position would be that of an astronomer who 

 was content with the truth arrived at by Kepler, that 

 the observed successive positions of the planets are due 

 to their paths being elliptical. Having ascertained the 

 nature of the planets' orbits, he would be justified in 

 NO. 1484. VOL. 57] 



asserting that the observed positions of the planets were 

 due to — that is, were caused by — the fact that their paths 

 are elliptical. But this would not be a sufficient causal 

 explanation of the planetary movements. There is 

 clearly a further question as to why the paths are 

 elliptical, and the elucidation of this question was reserved 

 for Newton. Hertwig would suggest that embryological 

 inquiry should stop short at a point analogous to that 

 gained by Kepler, and that we should content ourselves 

 with the assertion that the states which we observe in 

 individual ontogenies are what they are because the 

 organisms in question describe a sort of normal curve in 

 the courses of their development. It is hardly possible 

 to refuse one's sympathy to Roux when he declines to be 

 content to stop at this point, and urges that the know- 

 ledge hitherto acquired is but a preliminary to further 

 inquiry. Everybody who has studied and reflected upon 

 the facts of embryology must have felt the necessity for 

 further enlightenment as to why, and in virtue of what 

 inherent energies the ovum is able to go through the 

 complex succession of changes which lead to the 

 establishment of the adult individual. Various theoretical 

 solutions of the problem have been offered, but they 

 have not proved satisfactory. Roux steps forward and 

 shows that the only possible solution is by the method of 

 experimental investigation. Since he himself admits 

 that the problem was present to the mind of von Bar, it 

 is clear that his aim is not new, and in this unimportant 

 matter Hertwig is right ; but if the aim is not new, it 

 has only recently become practical, and Roux may lay 

 claim to the chief credit of having seen that the time was 

 ripe for trying to realise it. 



But it is one thing to have a legitimate and definite 

 object in view, another thing to devise the most appro- 

 priate means of attaining to it. Roux has entire faith in 

 experiment. Hertwig objects to the experimental method, 

 because in the act of making an experiment one disturbs 

 the normal course of vital phenomena, and obtains 

 abnormal results, from which nothing can with certainty 

 be predicated as regards natural processes. Butschli 

 has in a similar sense objected that the introduction of 

 disturbing factors into ontogeny involves a complication 

 in the results, which can only be justly estimated when 

 the elements of the mechanics of normal developmental 

 processes are well ascertained. The answer to this is 

 that no progress is possible if one allows one's self to be 

 discouraged by a priori objections and difficulties, and 

 that the method of experiment, so far as it has gone, has 

 been successful almost beyond anticipation. 



As regards the title "Developmental Mechanics" 

 {Entwickluftgsfnechanik), which Roux justifies at some 

 length, it need only be said here that the equivalent 

 "Experimental Embryology" most generally used in 

 England and America, though not expressly disavowed 

 by him, differs in its connotation from the title which he 

 has selected. Thus on p. 176, " Entwicklungsmechanik 

 bedeutet also die Lehre von den Entwicklungs- 

 bewegungen " : the essential idea is not contained in 

 the term Experimental Embryology. 



Roux's last task is to defend his practical methods and 

 results against the criticisms of Hertwig, who has not 

 hesitated to say that his preparations were so imperfect 

 in point of histological technique that nothing could be 



