THE ORIGIN OF NEW HEREDITARY CHARACTERS 



447 



continuity of the germ plasm and of the apartness of the germ plasm, 

 seemed to make the inheritance of somatic characters logically impos- 

 sible, for it seems that inheritance is possible only through linear cellular 

 descent, and germ cells are not the descendants of somatic cells, but 

 only of germ cells. Since Weismann's time expert opinion has been 

 prevaihngly anti-Lamarckian, though there have always been some 

 ardent Lamarckians in the field. The history of Lamarckian opinion 

 is an interesting one. Every year or so some experimenter comes for- 

 ward with an account of a series of experiments which he claims are 

 favorable to or prove the possibility of the inheritance of acquired 

 characters. During the following year or two some other investigator 

 repeats the same experiment with negative results or else points out 

 some logical or technical weakness or oversight in the experiment that 

 serves to invahdate it as proof of the point at issue. This situation has 

 been repeated so often that a justifiable skepticism is entertained about 

 every new report. All of us would like to beHeve in these experi- 

 ments, but experience has made us cautious. The last three years in 

 America have seen the publication of several very striking investiga- 

 tions, all of which seem to favor the Lamarckian hypothesis. For a 

 time opinion swung markedly toward the Lamarckian side, but just 

 at present, now that the newest experiments have been shown to be 

 either incapable of repetition or capable of being interpreted more 

 satisfactorily on other than Lamarckian grounds, the general pro- 

 fessional sentiment has once more become skeptical. Most of the 

 present-day students of organic evolution take the position that the 

 burden of proof is upon the Lamarckian, that such proof would be 

 welcome and most interesting, but that there is at present little hope 

 that such proof will be forthcoming. This, in brief, is the present 

 status of the problem. In the next chapter the whole question will 

 be thrashed out in detail. 



The mutation hypothesis. — The term "mutation" has come to 

 be appHed to a heterogeneous assemblage of different phenomena, 

 whose only common feature is that they involve changes directly in 

 the germ cells. At least four different kinds of phenomena are today 

 included under the one term: changes due to chromosomal aberra- 

 tions, gene or factor changes, chromosomal recombinations and segre- 

 gations involving lethal factors, and inductions or more or less 

 permanent injuries to the whole germ cell. Botanical geneticists' are 

 inclined to consider chromosomal aberrations as the real mutations (or 

 at least the most important mutations) because this type of germinal 



