146 
H. H. BARTLETT 
tions, with a frequency of approximately one mutation for every 400 
seeds sown. Correlated with the shape of the leaves are other char- 
acters, involving the size and branching of the plant and the pubescence 
and dehiscence of the calyx. The entire group of characters are co- 
herent ; it may be predicted from the shape of the seedling leaves alone 
that the other characters will appear in the mature plant. Never- 
theless it is quite impossible to imagine any necessary inter-dependence 
between the characters which cohere in this mutation. No systematist 
who did not know the parentage of mut. nummularia could possibly 
decide which of two dozen elementary species in my garden had given 
rise to it. 
Reciprocal hybrids between the parent species and the mutation 
appear to be strictly matroclinic, but the plants are still very young 
seedlings. The progeny of the cross mut. nummulariaX f. typica are 
all mut. nummularia, conversely, the progeny of the reciprocal cross 
are all/, typica, except for the fact that mut. nummularia appears with 
its usual frequency of one plant in several hundred. It seems that 
only female gametes bear the group of characters which distinguish 
mut. nummularia from /. typica. 
The other mutation which I wish to mention is Oe. stenomeres mut. 
lasiopetala. It differs from its parent species in a group of coherent 
characters, one of which is the hairiness of the petals. The solitary 
primary mutation when self-pollinated gave rise to a progeny consisting 
of typical Oe. stenomeres and mut. lasiopetala in a ratio suggesting 
1:1, although the former was in excess. It is highly improbable that 
a Mendelian explanation will apply to this case, either, but I do wish 
to "point out that such an explanation would necessitate viewing mut. 
lasiopetala as a dominant. A recessive could not have thrown the 
dominant parent. 
As my experience with this highly interesting group of plants in- 
creases I am more and more convinced that de Vries' conception of the 
origin of species is the true one. He believes that new species, differing 
from the old ones in a coherent group of characters, may come into 
existance at one step, by mutation. The evidence for this special 
view of mutation has been doubted by several critics, who have 
brought forward several destructive arguments. I believe that all of 
these arguments can be met. 
There is first the argument that Oenothera Lamarckiana is known 
only in horticulture, and may be a garden product; consequently that 
