18 HEREDITY, AND THE ORIGIN OF SPECIES. 
thousand plants. The evening-primrose of the Adiron- 
dacks and northern New England, Oenothera cruciata, has 
been found to give atypic individuals conforming to a 
single type, which is also represented by specimens that 
have been collected in a wild state, so that here also, we 
have the survival of a species which is still arriving in a 
large proportion of the progeny. The great - flowered 
evening-primrose, Oenothera grandiflora of the southern 
states has been grown during two generations and it also 
is found to give derivatives, one or more of which appear 
to be already represented in the flora of the region. 
One of the most interesting correlations to be made 
from a study of the results of these observations is to be 
found in the parallel mutations exhibited by the several 
species, in apparent contradiction of the principle of ra- 
diate variation and mutation (allseitige Mutationen). 
Among these are to be mentioned the origination of a form 
with cruciate flowers of the same general form as the 
Oenothera cruciata of the Adirondacks, from the species 
known as QO. biennis in Holland, which is not identical 
with any species known to grow wild in America. The 
same species has also been seen to give off a mutant having 
the character corresponding to the mutant nanella, coming 
from Lamarckiana, according to De Vries. Then an even- 
ing-primrose of unknown identity has been found on Long 
Island by Dr. Shull, far removed from the locality in- 
habited by any other cruciate plant, and strongly sugges- 
tive of a mutation from biennis or some other form native 
to that locality. Such facts merely show that the forms 
borne by nearly related species lie well within the limits 
of the morphological possibilities in saltations and that 
they may be expected to be duplicated in other observa- 
tions. 
Scattered through the literature of botany and horti- 
culture of the last century are scores of records of the 
