74 



THE AMEBIC AN NATURALIST [Vol. XLIV 



parently new characters may be formed by the gradual 

 dropping of these cumulative factors without any addi- 

 tional hypothesis. For example, in Nicotiana tabacum 

 varieties there is every gradation of loss of leaf surface 

 near the base of the sessile leaf, until in N. tabacum fruti- 

 cosa the leaf is only one step removed from a petioled 

 condition. If this step should occur the new plant would 

 almost certainly be called a new species ; yet it is only one 

 degree further in a definite series of loss gradations that 

 have already taken place. If it should be assumed that 

 in other instances slight qualitative as well as quantita- 

 tive changes take place as units are added, then it becomes 

 very easy, theoretically, to account for quite different 

 characters in the individual homozygous for presence of 

 all dominant units, and in the individual in which they 

 are all absent. 



Unfortunately for these conceptions, although I feel it 

 extremely probable that variations in some characters 

 that seem to be continuous will prove to be combinations 

 of segregating characters, it is exceedingly difficult to 

 demonstrate the matter beyond a reasonable doubt. As 

 an illustration of the difficulties involved in the analysis 

 of pedigree cultures embracing such characters, I wish 

 to discuss some data regarding the inheritance of the 

 number of rows of kernels on the maize cob. 



The maize ear may be regarded as a fusion of four or 

 more spikes, each joint of the rachis bearing two spikelets. 

 The rows are, therefore, distinctly paired, and no case is 

 known where one of the pair has been aborted. This is a 

 peculiar fact when we consider the great number of odd 

 kinds of variations that occur in nature. The number of 

 rows per cob has been considered to belong to continuous 

 variations by DeVries, and a glance at the progeny from 

 the seeds of a single selfed ear as shown in Table V seems 

 to confirm this view. 



There is considerable evidence, however, that this char- 

 acter is made up of a series of cumulative units, inde- 



