16 INTKODUCTION. 



growing organ, without exception, must, in a certain sense, be 

 capable of becoming modified in any such various directions as 

 are indicated by the common properties of the Hving substance 

 of the protoplasm. The abstract and paradoxical formula for 

 this position might be put thus : — Every living organ may, by 

 virtue of the properties inherent in its living cells, become any 

 other organ. 



Let us now return to our starting-point. We have seen 

 that every character of adaptation must be in a certain sense 

 hereditaiy ; for if those individuals of a species which first 

 acquired any given character by adaptation were incapable 

 of transmitting it to posterity as a part of their inherited 

 nature — particularly if the exciting causes were to be re- 

 moved — every newly acquired character would presently be lost. 

 The inheritability of this newly acquired character might also 

 be greatly increased if, for instance, it were transmitted through 

 a long series of varieties or species, while at the same time it 

 was imdergoing modification. This might occur if, from the 

 first, its fundamental character were such as must induce 

 specialisation of its function. Now we have seen that even so 

 highly specialised an organ as a gill is, or seems to be, must 

 yet be capable of numerous modifications ; for its primary use 

 depended partly on other functions which were capable of 

 further modification, i.e. specialisation, in a mode analogous 

 to that by which the gill became an organ of respiration. And 

 the more manifold the independent and latent properties are of 

 a newly constituted organ of adaptation, the greater will the 

 probability be that it will be transmitted by inheritance to all 

 the divergent descendants of the parent form, and be at the same 

 time modified to meet their altered functional requirements. 

 But the more specialised an organ is — that is to say, the more 

 one single purpose is developed to the prejudice of the latent 

 functions — the harder will it be for it to adapt itself to new pur- 

 poses, and so it will probably be transmitted to the descendants 

 of the parent form but little altered. Hence it is impossible to 

 establish an ct priori distinction between Characters of Adapta- 

 tion and Characters of Inheritance, and we perceive that most, 

 perhaps all, of the characters now in a great measure hereditary 



