260 



The Journal of Heredity 



will justify the enforced celibacy of so 

 large a body of superior young women 

 as is now engaged in school teaching. 



The magnitude of the problem is not 

 always realized. In 1914 the commis- 

 sioner of education reported that there 

 were, in the United States, 169,029 

 men and 537,123 women engaged in 



teaching. Not less than half a mil- 

 lion women, therefore, are potentially 

 affected by the institution of peda- 

 gogical celibacy — an institution which 

 is to be compared with that of sacer- 

 dotal celibacy in the amount of perma- 

 nent harm that it is capable of doing 

 to the race. 



The Status of the Presence and Absence Hypothesis 



To explain the inheritance of genetic 

 differences, we assiune the presence in 

 each germ-cell either of one of two 

 alternative factors (Morgan and his 

 school), or the presence or absence of 

 one unpaired factor (Bateson and his 

 followers). Bateson's "presence and 

 absence hypothesis" applies to Men- 

 delian character differences if they are 

 quantitative, and is extended to the 

 genetic differences in the germ-cells, 

 whether these result quantitatively or 

 qualitatively. It demands, I think, the 

 following subsidiary hypotheses : 



1. That even qualitative differences 

 are due to presence and absence, and 

 not to substitution of factors. 



2. That dominance (or prevalence) 

 is always the mark of the presence of a 

 factor. 



3. That recessive factors do not exist. 



4. That even where dominance or 

 prevalence is absent, the side on which 

 the factor is present can safely be 

 guessed at. 



5. That all cases of multi])le allelo- 

 morphs (triangles of Baurj are due to 

 complete linkage. 



6. That double presence of a domi- 

 nant (or prev^alent) factor is often more 

 effective than its single presence. 



To infer the nature of the germinal 

 factors from the results of their opera- 

 tion is not usually possil)le in such com- 

 plicated biochemical machines as plants 

 and animals. Hence the comparative 



simplicity of the presence and absence 

 hypothesis is, I think, illusory. In 

 working with a quantitative Mendelian 

 difference it is permissible, as a mathe- 

 matical convention, to regard the result 

 of one factor as zero;^ but this does not 

 apply to qualitative differences. 



1. The two-factor hypothesis escapes 

 the speculation that the dominance or 

 prevalence of a "character" informs us 

 as to the nature of the germinal factor 

 causing it. 



2. The two-factor hypothesis applies 

 equally to qualitative differences: e.g., 

 peloric and regular flowers. 



3. The two-factor h\'pothesis is con- 

 firmed by every case of multiple allelo- 

 morphs discovered. 



4. Fluctuation of dominance and 

 absence of dominance are, I think, 

 more readily intelligible on this hypo- 

 thesis. 



5. Few or no subsidiary hypotheses 

 are needed. 



6. Dominant mutations, which are a 

 stumbling-block to users of the one- 

 factor hypothesis, agree about as well 

 as recessive mutations with the two- 

 factor hypothesis. 



As Morgan and Sturtevant have 

 already shown, the two-factor hypo- 

 thesis can be used in Mendelian work 

 as well as or better than the presence 

 and absence hypothesis. But in vari- 

 ous cases both may be true. 



John Belling. 



A New Strain of Hybrid Sheep 



A strain of Ham])shirc(l()wn-Ram- 

 bouillct hybrid sheep has V)een almost 

 perfected at the New Hampshire li^xperi- 

 ment Station, according to Dr. Charles 

 B . Davenport , who started it . The new 



breed is said to show in a high degree a 

 union of esi)ecially valuable qualities of 

 fine wool and good conlVirmation. A sec- 

 ond hybrid generation of a Southdown- 

 Rambouillet cross has been produced. 



'See "Inheritance of Length of Pod," by John BcUing, in Journal of Agricultural Research, 

 December, 1915. 



