446 THE AMERICAN NATURALIST [Vol. LV 



three seed patterns are found in association with dark- 

 blue flowers in different pedigrees. In the second place 

 there is the case of the dark-blue plant, VIII-27, which 

 produced seeds with the light color characteristic of light 

 blues. 



In the great majority of mutations described up to 

 the present time the new character is recessive to the nor- 

 mal one (1). They are apparently due to the loss or 

 inhibition of a previously existing factor. It is interest- 

 ing to note that the two factors for light blue and striped 

 white here reported are both dominant, the former com- 

 pletely and the latter incompletely so. Furthermore, 

 many of the other variations observed are in the nature 

 of additions. Yellow color on the front of the banner 

 and orange lateral spots, although their heredity is not 

 yet known, are certainly to be considered as in the nature 

 of additions. 



The mutations in Lupinus represent three categories 

 (11). In the whites a positive character has been lost. In 

 the light blues and striped whites a character has been 

 replaced by another. This would naturally be supposed 

 to be due to some alteration of the factor governing the 

 somatic character. In the case of the orange spots and 

 lemon-yellow fronts one is led to suppose the addition 

 of a new factor, especially so in view of the fact that they 

 are not constantly associated with one another or with 

 white flowers. White flowers may occur without either 

 lemon fronts or orange spots or with both or with one 

 and not the other. 



These three categories of characters naturally suggest 

 that there are also three sorts of factorial bases for them. 

 Loss of a character appears readily explicable either on 

 the assumption of an actual loss of the factor in the chro- 

 mosome or of its becoming latent. Multiple allelomor- 

 phic factors seem to be located according to the work of 

 Morgan and his associates (8) at identically the same 

 loci in their respective chromosomes and must therefore 

 be thought of as different changes of the same factor. 



