to be guided by symptoms of a particular tendency to variability, and hence chose by preference seed from plants with fasciated stems, split leaves or other variations. I also sowed, as far as room permitted, seed of annual garden plants, bought in shops.
It is clear that, notwithstanding the immense amount of work involved, the chances of success were exceedingly small. Yet I was lucky enough to find the very thing wanted. Among a hundred species there was a single one which proved to be mutable, at first but in a small degree, but sufficiently to decide me to abandon nearly all other experiments and to study this one plant as thoroughly as possible. Of the other species I had in the meantime obtained a number of monstrous races; these I continued to cultivate, but not the others.
The plant referred to was Oenothera Lamarckiana, a species of American origin, which has here escaped from cultivation, as did formerly both the evening primroses, Oenothera biennis and Oenothera muricata, which at the present time are quite common on our sand dunes. Oenothera Lamarckiana, the large evening primrose, surpasses both other "species in size of flowers, but for the rest is very much like them. This plant was first described by Lamarck as Oenothera grandiflora, but by this name a number of other species of Oenothera are known. Seringe changed the name to Oenothera Lamarckiana, which name has been retained.
In 1886 I collected a quantity of seed from wild plants of Oenothera Lamarckiana and also transported a number of rosettes of biennial specimens to the botanical garden at Amsterdam. The next year they flowered profusely and produced a large quantity of seed.
The seed obtained from wild plants was sown in 1887 and yielded at once what we desired. For among the plants obtained from it, there were three which, though agreeing among each other, possessed characters entirely deviating from those of the rest. This species was, therefore, able to produce at least one new form. The new form differed more from the mother species than the three species above mentioned did among each other. The leaves were broader, rounder and more obtuse, the buds swollen and the fruits small. The stems were small, weak and remained brittle even in autumn. At the tips of the branches the young leaves and buds were collected in crowded rosettes, so that at first the plants were denoted 'roundheads.' In a number of other particulars they differed more or less from the ordinary form, in fact they did not entirely agree with it in a single point. The most important difference, however, was the inability to produce good pollen. The anthers of the mother species are, when open, thickly covered with a sticky powder, which is entirely lacking in the case of the roundheads. The anthers of the new form are dry, what little pollen there is is shriveled, for the greater part unfertile and entirely unfit for