NOTES ON THE ORIGIN OF THE MOSS-ROSE. 
39 
Genetic Significance. 
In concluding this inquiry, it may be of interest to add a few genetic 
notes on the probable nature and significance of the three definite 
appearances of this " specific " bud-mutation, after more than 2,000 
years of intensive cultivation and vegetative propagation. 
That the ' Moss ' character in Rosa is a genuine bud-mutation, 
and not a fluctuating variation, or bud-variant, is evident from its 
somatic persistence through many bud-generations, and its germinal 
persistence through various seed-generations. 
Rivers (1840) states that : " Plants produced by the seed of the 
Moss-Rose do not always show Moss ; perhaps not more than two plants 
out of three will be mossy, as I have often proved." 
Again, Darwin (1893) states that : " Mr. Rivers informs me that 
his seedlings from the old Single Moss-Rose almost always produce 
Moss-Roses.' ' 
We do not yet know definitely whether the ' Moss ' character 
in Rosa is to be identified with a single Mendelian factor or not, though 
all the evidence so far is in favour of its being a simple dominant. 
If it is, and Rivers' matings of the Single Moss-Rose were, on the 
average, one-half selfings and one-half crossings with other Roses, 
as they probably were, judging by his methods, the Mendelian ex- 
pectation, on the average, would be the actual ratio he obtained, 
viz. 2 Moss ; 1 Plain. It is evident that ' Moss ' is a dominant 
character, for if it were recessive no hybrid Moss-Roses would have 
the ' Moss ' character, and we know that some have. It is also 
clear that the Single Moss, a bud-variation from the Old Moss- 
Rose, is a heterozygous dominant, for according to Rivers (1840) 
and Darwin (1893) it throws " Plain " as well as ' Moss ' Roses 
from seed in the proportion of about one in three. 
We have already seen that the Old Moss-Rose produced " plain " 
(i.e. " unmossed ") Cabbage-Roses. So that it is probable that 
the first mutation of the Moss-Rose was itself a heterozygous 
dominant for ' Moss.' 
In view of the important results recently reported by Morgan 
(1919) and his colleagues, in their experiments with Drosophila, which 
have led them to formulate the chromosome theory of heredity, it 
seems on this hypothesis that if the first mutation of the Moss-Rose 
was a heterozygous dominant, the mutational change would take 
place in one of the chromosomes in a single locus. In accordance 
with the " presence and absence " method, this mutational change 
in a single locus from m to M involves the presence of an additional 
factor M, which is dominant to the normal allelomorph m from which 
the factor M is absent. 
This conception, however, does not necessarily imply the actual 
presence or absence of a structural gene as Morgan seems to infer, 
and in the present state of knowledge it seems safer and sounder to 
continue the use of the non-committal term " factor " with its 
