Mutations and Evolution . 
178 
in which the origin of new divergent characters is now or has 
recently been taking place. 
Turning now to the recorded cases of mutation in the literature, 
they form such an array that it would be a colossal task to attempt 
even to enumerate them. The condition is so common in flowering 
plants that it may be considered exceptional to find a species 
without any mutational varieties. A few instances will be selected, 
almost at random, for mention here. First we will consider 
doubling. 
Doubling. 
The gardeners’ tradition that “doubling” or the transformation 
of stamens into petals,' is a result of cultivation has long been dis¬ 
proved. It is evident from the records that doubling of flowers 
occurs “ spontaneously ” both in wild and cultivated species. It is 
a typical mutation, although the completeness of the transformation 
of stamens into petals, and the associated changes in the flower, 
may vary in different cases. Breeding experiments have shown 
that the double condition is in some forms a Mendelian dominant, 
but more usually a recessive. Bateson (1909, Chap. XI) has 
considered the cases studied before 1909. The work of Miss 
Saunders on the inheritance of doubleness has been most ample, 
and some of her results may be summarized here. Taking the 
simpler cases first, she finds (1917) that in the Welsh poppy, 
Meconopsis canibrica, doubleness is dominant, the F 2 from a cross 
containing three doubles to one single. The doubleness results 
from a variable degree of petalody of the androecium and gyncecium. 
In the hollyhock (Althcea rosea and A . ficifolia) the F t is intermediate 
and the F 2 gives the 1 :2: 1 ratio, i.e., neither condition dominant; 
in carnations (Dianthus caryophyllus) doubleness is a simple 
dominant, 2 while in sweet william (D. bavbatus) singleness is 
dominant and the P 2 from a cross contains three singles to one 
double. 
In Chelidonium majus Sax (1918) finds that doubleness is 
apparently a simple recessive, but the P 2 gives a continuous series 
from singles to full doubles. Statistically considered, there is much 
1 Doubling may also occur, as in Fuchsia, by an increase in the number 
of petals without any change in the stamens, or it may result from change of 
carpels to petals, or from all these conditions combined. 
s According to the breeding work of Norton and of Batchelor the doubleness 
of carnations appears to be of two distinct types or degrees, full doubles and 
standard doubles, the latter having some stamens. Full double x single 
gives an F, which is standard double, and an F, with the three types in the 
ratio 1:2: 1. Miss Saunders finds, however, all degrees of doubling in the 
DD’s and DR’s, which are therefore indistinguishable. 
