quantity among 50,000 other plants. To this it may be replied that the phenomenon of fluctuating variation, as it appears in Œnothera, has been studied in detail by de Vries and has been exactly determined both for the central species and for the different subspecies (mutations). In all of them it occurs on a large scale, but not one of the specimens above mentioned belongs to it. These 800 have very special characteristics, by which they can be sharply distinguished from the fluctuating variations. And, as is especially remarkable, they are not in every respect different from each other, but may be arranged in seven natural groups, each of which possesses exactly the same systematic value as that particular combination of specific characteristics to which the name of Œnothera Lamarckiana has been applied.
The number of individuals of those seven groups, of which de Vries has observed the spontaneous appearance, is, however, most unequal and varies between 1 (Œnothera gigas), 56 (Œnothera albida), 350 (Œnothera oblonga), 32 (Œnothera rubrinervis), 150 (Œnothera nanella), 221 (Œnothera lata) and 8 (Œenothera scintillans).
De Vries has studied the mutations thus arising, some of which are rare and some more common, with the utmost care, and has followed them during the whole of their existence.
Œnothera gigas, which has only once arisen in the Lamarckiana group and which is characterized by much larger-sized flowers, a different shape of the leaves that form a rosette at the root, more thickly set leaves along the stem, etc., was sown by de Vries in 1897, after he had obtained seed, thanks to artificial fecundation, the possibility of self-fertilization being excluded. He thus obtained 450 plants, which, with the exception of a single one, exhibit all the characteristics of the Œnothera gigas with perfect constancy. The exception was not a retrogression towards O. Lamarckiana, but a new deviation, provisionally indicated as O. gigas nanella. From 1898 to 1900 further propagation by seed has been effected during three generations and under very strict precautions; and until now all the descendants of that one mutation of 1895 remain perfectly constant; de Vries has actually seen the species O. gigas come forth out of O. Lamarckiana, first in nature, afterwards in his own nursery-garden. It appears to be a very strong plant, which, if it had to fight for its existence against O. Lamarckiana, in equal numbers and under the same circumstances, would probably prove to be the winner.
The second mutation, Œnothera albida, which occurred 56 times during the experiments, shows another character. It is a feeble plant, and was originally considered a pathological deviation, which, however, in the later generations has proved itself to be none the less constant, and, though but little fertile, produced 86 plants in 1898 and 36 in 1899.