220 



THE AMERICAN NATURALIST [Vol. XLVI1 



anthers in a flower containing 8 or 10? Not because it is 

 absent from the other anthers, because the pollen from 

 these anthers transmits the character. It is not a ques- 

 tion of segregation then, but one of environment. 



Evidently the gene is inactive or latent, for we know 

 there is something present which for convenience we call 

 a gene, and yet we can not see any of the visible signs of 

 its presence, such as we see in the affected anthers. On 

 a morphological conception it must be there ; physiolog- 

 ically for the time being, so far as we can determine, it is 

 non-existent. The inactivity we may suppose is due to a 

 lack of a properly adjusted environment. This proper 

 adjustment is only true of a few anthers in the F x plants. 

 We believe this scarcity to be due to two kinds of latency 

 — inactivity of the gene as in the pure abnormal and inac- 

 tivity of the gene because of association with the cell ma- 

 terials that trace their lineage back to the sperm of the 

 normal father. But latency is a vague term. In ge- 

 netics, it is used to describe the period between the disap- 

 pearance of a character and its reappearance. By push- 

 ing this conception to its logical conclusion it is clear 

 that one can practically never prove the origin of a new 

 character. Fasciation, while new to Nicotiana, is phy- 

 logenetically an old character. The production of purple 

 fruits in Rosa would mean, phylogenetirally, the reap- 

 pearance of a latent character, for purple fruits are com- 

 mon to the Amelanchiers and to a species of Pyrus. zl 



The characters of the whole plant kingdom would be 

 in a state of latency and patency, of inactivity and ac- 

 tivity. To determine whether a character were new or 

 not would involve a canvass of that part of the plant 

 kingdom phylogenetically older than the family under in- 

 vestigation. Of course, we speak of segregation in 

 phylogenetical lines, but the term has a different mean- 

 ing in such cases. My F 2 normal segregates are pure 

 and will breed true for absence of abnormalness, I be- 

 lieve, for any number of generations unless a new muta- 



81 Pyrus Niedwetzkyana. 



