142 
R. Ruggles Gates. 
the type of biennis only in having pale yellow petals. It is found 
occasionally on the sand dunes and must have originated as a 
mutation. 
The well-known uniformity and constancy of (E. biennis , 
together with the apparent impossibility of crossing in the regions 
from which these plants were derived, greatly enhances the 
importance of the mutations obtained from them. This species, the 
only one with which Linnaeus was at first acquainted, was originally 
brought from North America in 1614—the first evening primrose 
to be brought to Europe. 1 It was soon naturalized on the Dutch 
sand dunes, where it has retained its characters with remarkable 
uniformity for three centuries until the present time. Davis (1914) 
admits that no species in cultivation has greater claims to be 
regarded as a pure species, and at one time was willing to admit 
that if mutations were produced by it they would go far to prove 
the correctness of the mutation hypothesis that the production of 
new forms is not merely a phenomenon of hybridity. Indeed it 
would be difficult to find any species whose claims to “purity” or 
freedom from crossing are better authenticated. Not only is it 
self-pollinated, so that crossing would be a rare event at best, but 
for three centuries in these Dutch localities it has been more or 
less completely isolated from its American relatives. Even if its 
ancestors were crossed before they left America, 300 generations 
of self-fertilization would be more than ample to reduce it to a 
homozygous condition again, 2 unless a condition of balanced lethal 
factors maintained it in a heterozygous state. 
But the Drosophila work shows (Muller, 1918)) that such a 
condition of balanced lethals is not the result of crossing, but of 
alterations which arise in the germ-plasm. Another fact against 
the hybridity theory is this. The mutant sulfurea , which has been 
shown to arise occasionally in cultures, is also “far from rare ” on 
the Dutch dunes (de Vries, 1915b). Apparently it was first 
mentioned in 1687 by Hermannus in Hort. Acad. Lugduno-Batavi 
1 For the early history of some of these forms see Mut. Factor , p. 47. 
2 Without dealing with the mathematical formulae for inbreeding involved 
(see East and Jones, 1919, p. 91), it may be pointed out that even if a large 
number of independent differences were involved in an original cross, the 
population from self-pollination would be practically all homozygous by the 
tenth generation. The population would then be expected to contain a 
number of different homozygous types. It might be anticipated that the 
weaker of such types would fail in the struggle for existence. But if we 
examine the facts, we find that CE. biennis in Holland is remarkably uniform 
an 1 has always been so, never showing any variations on the dunes except 
the sulfurea mutation, and leptomeres with cruciate petals. There is thus a 
complete absence of data on which to found a hypothesis of hybridity, while 
the evidence to the contrary is strong. 
