3 2o AUGUST WE ISM ANN [april 



Certain cells of the developing animal contain, to begin with, many 

 " Anlagen," and which of these shall become active is decided by the 

 position the cells take up in the organism, i.e. through the influences 

 exerted upon them by the rest of the living parts of the animal. 



All these phenomena, and also more especially that which Koux 

 has called " post-generation," make it plain that my theory of ontogeny 

 requires some modification. When it was put forward, two funda- 

 mental postulates for the explanation of ontogenetic differentiation 

 occurred to me as possible — either that there is a regulated progressive 

 breaking up of the " Anlagen " — masses contained in the germ-plasm 

 into smaller and smaller groups ; or that the collective " Anlagen " 

 remain together in all the cells of the organism, but that each of them 

 was responsive to a specific stimulus, which alone could excite it to 

 activity ; the former view may be described as a theory of dispersal, 

 the latter as one of liberation. I inclined to the former, because, in 

 face of the facts before us at that time, it seemed on the whole the 

 more probable ; and because I found myself in the position of being 

 obliged to make a choice if I wished the applicability of my principles 

 of heredity to be tested in relation to all aspects of the problem. I 

 found myself in the position which the present Prime Minister of 

 England, Lord Salisbury, 1 referred to in his Oxford address as one in 

 which politicians found themselves often, but naturalists never — that 

 of having to choose the lesser of two evils or uncertainties. The mere 

 observer of facts certainly does not find himself in this position, but 

 it is otherwise with the theorist who is confronted with the task 

 of associating in idea the phenomena of a whole domain, filling 

 up the gaps in the knowledge of the moment with probabilities, 

 seizing the essentials of both, and combining them on a definite 

 principle into a provisional theory. He must choose between 

 greater and lesser probabilities if his theory is to be worked out 

 at all, just as the politician must act, if the whole machinery of the 

 State is not to come to a standstill. 



This is by no means meant to indicate that I now consider my 

 idea of a regular dispersal of the germ-plasm in the course of ontogeny 

 incorrect ; it rather appears to me a more indispensable assumption 

 than ever, but I recognise that liberation (Auslosung) plays a larger 

 part in ontogeny than I had hitherto believed, and that the assumed 

 breaking up of the germ-plasm does not, at least in many cases, take 

 place at the first division of the ovum, but later. 



It is not my intention to enter more fully into this question here, 

 but I should like to emphasise the fact that a modification of my 

 theory of ontogeny can in no way react upon the theory of heredity 

 in the strict sense. My ideas as to the mechanism of heredity would 

 remain unaltered even were I to adopt a pure liberation-theory such as 



1 "Address by the Most Hon. the Marquis of Salisbury," Report Brit. Assoc. Oxford, 

 1894. 



