ARE UNIT-CHARACTERS CONSTANT ? 183 



breed true, as blue pigeons do; a race of barred Pl^-mouth 

 Rock fowls of the same color in both sexes. These ideals the 

 student of genetics says are unattainable and he can gi\'e good 

 reasons for so regarding them. Nevertheless breeders will 

 doubtless continue to try for them and it is hardly safe to say 

 that success is impossible. Most advances in practical affairs 

 are made by those who have the courage to attempt what 

 others with good reason think unattainable. When such 

 attempts have succeeded, the world simply revises its classifi- 

 cation of things attainable and unattainable, and makes a 

 fresh start. 



Many students of genetics at present regard unit-characters 

 as unchangeable. They consider them as impossible of modi- 

 fication as are the atoms. To recall Bateson's comparison, 

 the carbon and oxygen of carbon monoxide, CO, are each 

 unchangeable. Adding another atom of oxygen does not 

 alter them, though it changes radically the compound fonned 

 which becomes carbon dioxide, CO2, possessed of very differ- 

 ent properties. But the carbon and the oxygen are still there, 

 unaltered and recoverable. This question is one of great 

 practical importance, — are unit-characters as constant as 

 atoms, so that we can merely recombine them, or are they 

 different in nature from atoms so that we can modify as well 

 as recombine them. Much careful work has been devoted to 

 the solution of this question. It was at first assumed from 

 chemical analogy that characters which behave as units in 

 heredity must, like C and O in the case of carbon dioxide, 

 emerge from combinations unmodified. But presently case 

 after case came to light in which this was not true. Albin- 

 ism emerged from crosses tainted with color; clear yellows 

 emerged from crosses intensified to red, or diluted to cream, 

 or sooty with minute quantities of black; patterns such as 

 are seen in Dutch or in English rabbits, or in hooded rats, 

 emerged considerably altered in appearance. Facts such as 

 these were interpreted in two different ways. It was as- 

 sumed by some that the actual unit-character, factor, or 

 gene involved was subject to quantitative and possibly to 



