322 



THE AMERICAN NATURALIST [Vol. XLIII 



to the mass of statistical data involved. For the evidence 

 of the statements here made the reader is referred to 

 these. 



Unicellular organisms are essentially free germ cells 

 —germ cells that are subjected to the immediate action 

 of the environment, both direct and selective. For long 

 periods they propagate without that intercrossing which 

 so tremendously complicates the study of heredity in 

 higher animals. Here if anywhere we should see readily 

 the effects of environment and of selection in modifying 

 a race. 



Let us look first at the direct action of the environ- 

 ment: the "inheritance of acquired characters." It has 

 commonly been thought that under the conditions found 

 in these organisms "acquired characters" are readily in- 

 herited. This is because the progeny arise by division 

 of the parents ; they are therefore the same as the parent. 

 It would seem a matter of course therefore that they 

 should have the same characteristics as the parent, how- 

 ever these characteristics arose. 



But when we examine just what occurs in the produc- 

 tion of the new individuals, we find— as usually happens 

 when we look closely at biological processes— that the 

 thing is not so simple after all. We find that in repro- 

 duction the characteristic features of the parent disap- 

 pear 3 and are produced anew in the offspring. Thus in 

 Paramecium (Fig. 1) the characteristic form of the ends, 

 the oral groove, the shape of the body— these disappear 

 during fission, and reappear in the growth of the young. 

 In Stylonychia (Fig. 2) all the appendages are absorbed 

 at division; they appear anew in the young, in their 

 characteristic structure, number and distribution, by a 

 process comparable to the development of organs in a 

 higher animal. 



3 Certain exceptions to this, of no theoretical importance, are mentioned 

 in the original papers. Certain characters sometimes pass directly to one 

 of the offspring, but their multiplication is of course always by new 



