424 ORIGINATIVE FACTORS IN EVOLUTION: 



5. Correlation of Variations. 



The tendency of modern research has been to lay emphasis 

 on the idea of hereditary particulateness, that the character- 

 istics of organisms are made up of elementary units, without 

 intergrades, as sharply separated from one another as the 

 chemical elements. This is the idea of " unit characters ", 

 independently heritable, and independently variable. It is 

 very striking that a trivial feature in the hands a reduc- 

 tion of the index and middle finger (in spite of the presence 

 of a little extra triangular bone at their bases), and a con- 

 sequent projection of the ring finger, should behave as a 

 Mendelian character for at least four generations and be 

 found in fifteen out of thirty-six descendants of the family 

 investigated. (See H. Drinkwater, Journ. Anat. Pliysiol., 

 L., 1916, pp. 177-186, 14 figs.) There is indirect evidence 

 that particular unit characters are represented by particular 

 particles (factors, determinants, or genes) in the germ-plasm, 

 or perhaps by ultra-microscopic differences of architecture, 

 and the idea works well, like the atomic theory in chem- 

 istry. But it has its limitations and it must not be pressed 

 so hard that we lose sight of the unity of the organism even 

 in the germ-cell phase of its being, and of the fruitful con- 

 ception of correlated variations. An exaggeration of the 

 idea of particulateness leads to a view which is too mechan- 

 ical to fit living creatures, as if the organism evolved like 

 a machine perfected piecemeal by the adding on of many 

 little patents independent of each other. A reaction may 

 be seen in the recent book by Prof. T. H. Morgan and others 

 on The Mechanism of Mendelian Inheritance (1915), 

 where it is insisted that the so-called unit character is only 

 the most obvious or most significant product of the postulated 



