366 EXPERIMENTAL STUDY OF INHERITANCE 



of the clearest and surest facts about an organism that it is a 

 unity ? It lives as a unity, does it not evolve as a unity ? 



We cannot here enter into a discussion of the alleged anthro- 

 pomorphism or sociomorphism of what we flatter ourselves by 

 calling " pure science." That is a very interesting thesis, and 

 worthy of much discussion. But we wish to refer for a moment 

 to the idea of the " piecemeal patenting theory " of evolution, 

 since it seems to us that the facts brought to light by Mendel 

 and the Mendelians are sufficient to show that there is some truth 

 in this way of looking at the organism. 



It has been shown that some organisms have clear-cut, we 

 may almost say crisp, unit characters, which behave in inherit- 

 ance as if they were independent constituents, being transmissible 

 en bloc and in their entirety not blending with analogous 

 characters, but remaining quite distinct, and developing in 

 absolute intactness and exclusiveness or not at all. 



The Mendelian facts, as Bateson says, lead us to regard the 

 organism as " a complex of characters, of which some at least 

 are dissociable and are capable of being replaced by others. . . . 

 We thus reach the conception of unit characters, which may 

 be rearranged in the formation of the reproductive cells. It 

 is hardly too much to say that the experiments which led to 

 this advance in knowledge are worthy to rank with those that 

 laid the foundation of the atomic laws of chemistry." 



Weismann has not paid much attention to Mendel's Law, 

 because he regards the basis of facts as still insufficiently broad, 

 and because he sees so many discrepancies in the experimental 

 results ; but it may be pointed out that the general idea of in- 

 dependently heritable unit characters is not inconsistent with, 

 but rather corroborates Weismann's picture of an inheritance as 

 composed of numerous sets of determinants or primary constitu- 

 ents, each corresponding to an independently variable and 

 heritable structure. It is quite possible that the germ-cells of 

 the hybrids of two distinctively contrasted parents do not separate 



