1896 ] ^7 [Conkliu. 



These evidences are, I know, not the strongest ones which can l)c 

 adduced in support of the Lamarckian factors. There are at present a 

 relatively small number of such arguments which seem to be valid and 

 the great force of which I fully admit. But the cases which I have 

 cited are, I believe, fair samples of the majority of the evidences so far 

 presented, and in the face of such "evidence" it is not surprising that 

 one who is himself a profound student of the subject and a convinced 

 Lamarckian praj'^s that the Lamarckian theory may be delivered from 

 its friends.* 



6. Another line of evidence, and by far the most promising, is that of 

 direct experiment. So far most of the experiments which have been 

 carried on to determine this question have been carried only half way 

 to a conclusion — they have shown that characters are acquired, they 

 have usually failed to show that they are transmitted to descendants. 

 Among animals one of the best known cases is the inheritance of 

 epilepsy and other disorders in Guinea pigs, due to certain nervous 

 lesions of the parents. But Romanes,! who spent much time in trying 

 to corroborate these results, concludes as follow^s : "On the whole, 

 then, as regards Brown-Sequard's experiments, it will be seen that I 

 have not been able to furnish any approach to a full corroboration." 



Among plants, on the other hand, there is more and better experi- 

 mental evidence, but it is not by any means as full or satisfactory as 

 could be wished. Of one thing we may be certain : a satisfactory solu- 

 tion of the problem can be reached only by experiment. The mere 

 observations and inductions of the morphologist, while affording valu- 

 able collateral evidence, can never furnish the crucial test. As long 

 as we deal mereh' with probabilities of a low order there will be pro- 

 found differences of opinion : e. g., Cope believes in all the Lamarckian 

 factors; Romanes rejects use and disuse, but believes in the others; 

 Weismann rejects all of them. Why? Is it because each does not know 

 the facts upon which the others build? Certainly not. Those so-called 

 facts are merely probabilities of a higher or lower order, and to one 

 man they seem more important than to another. 'No conviction based 

 even upon a high degree of probability can ever be reached in this way. 

 There is here a deadlock of opinion, each challenging the other to pro- 

 duce indubitable proof. This can never be furnished by observation 

 alone. Possibly even experiment may fail in it, but at least it is the 

 only hope. 



Conclusion. 



On the whole, then, I believe the facts which are at present at our 

 disposal justify a return to the position of Darwin. Neither Weismann- 

 ism nor Lamarckism alone can explain the causes of evolution. But 

 Darwinism can explain those causes. Darwin endeavored to show that 



* H. F. Osbom, Evolulion and Hrredily, Biological Lectures, 1890. 

 tG. J. Romaue-, Post- Darwinian Questions, 1893. 



