Some important consequences result, however, if systematic care is brought to bear on the matter. This tendency to variation in the second generation of crossed plants has often been noted, and it bears out very distinctly the conclusions to which Darwin came.

The hybridiser takes advantage of this variation, as others have done, to select some forms and rigidly suppress others, in order to obtain well-marked varieties of the plants he experiments with. In illustration, I may take the following from Rimpau's account of his experiments on crossing wheat: By crossing a white English long-eared, dense wheat, and celebrated as a heavy cropper, with a red, looser German wheat, remarkable for its resistance to winter cold, Rimpau hoped to obtain a variety uniting both the above qualities. As regards the property of resistance, he failed, and he eventually gave up the attempts in face of the advantages offered by the so-called Square-heads, which then came into the market. His experiments, even with the above varieties, are worth noting, however, for they show how promising the results of carefully conducted crossing and selection may be.

The crossing was done in 1875, in both directions. In 1876 the few grains obtained were found to yield plants almost all alike, with the long loose ear of the German parent, but the paler colour of the English wheat.

In 1877 the plants, obtained by sowing the finest grains, were found to consist of pure white, pure red, and of forms which appeared to vary and revert in all possible degrees as regards colour, density, and other characters intermediate between these.

By carefully separating the closest and densest white wheats from the closest and densest red ones, he got in 1878 a large number of each coming nearer to the type sown than did the mongrel forms intermingled with them: these reversions and intermediate forms were then rigidly eliminated, and only the deepest coloured and densest red and white forms again sown.

In 1879 these two chosen varieties were constant, so far as concerned those selected from the crossing of female English white with male German red wheat, and the following year proved the constancy of the red variety in the reciprocal cross. In 1886 all four varieties - i.e. the two reds and the two whites of both the crossings - had become constant.

Still more instructive are the results of the cross between the same white English non-bearded wheat and a red German bearded wheat.

The first results of the crossing in 1875 showed the loose ear of the German mother, but was paler in colour; while the influence of the English father was shown by the absence of beard.

From the reversions and mixtures of the mongrels showing reminiscences of the parents in all degrees in 1877, rigid selections and re-sowings were made as before, and Rimpau eventually got four very distinct varieties, two red and two white, a bearded and a beardless form of each, and these were declared fixed and constant in 1879 - 1882.

Passing over many similar results, and merely noting a very successful variety got from a cross between a very early ripening loose red American wheat and the dense heavy cropping English Square-head - the crossed variety which has proved very suitable for certain light soils and dry climates on the Continent, which demand very rapid ripening, and are therefore of great physiological and technical interest - I must pass on to note the curious result of the successful hybridisation of wheat and rye. This cross has been effected several times, and first in this country according to reports from Edinburgh (1875), New York (1886), and elsewhere, and Rimpau's careful experiments seem to leave no doubt on the matter.

First I must remind you that wheat (Triticum) differs from rye (Secale) in several marked characters, such as the breadth and shape of the glumes, the number of flowers in the spikelet, etc.; and that the cultivated rye differs from cultivated wheats in the characters of the straw, in having long ears, and in its flowering glumes remaining widely divaricated for some days when in flower.

In 1888 Rimpau removed the young stamens from the German wheat referred to, and pollinated the stigmas with pollen from a long-eared rye. Four sound grains were obtained, looking like wheat-grains.

The history of one of these grains was as follows: In 1889 it yielded ears which were peculiarly narrow and long, and its stalks were also much longer than the wheat: the flowers remained exposed, with widely open paleae, for several days, and the grains were very peculiar, though wheat-like.

Fifteen of the best grains were selected, and in 1890 three of the resulting plants proved to be a wheat of the Square-head type and one quite sterile. The others retained the elongated, narrow, brownish-red ears, the flowering glumes again opening wide for some days. This last is a characteristic of rye, but not of wheat.

A long series of natural hybrids of wheat, barley, and oats are also described and discussed by Rimpau, as well as artificial crosses - some very remarkable - of barleys, but they must be passed over here.

Peas rarely become hybridised naturally. According to Darwin, H. Miiller, and Focke, the flowers are little visited by insects in our countries, though the mechanism points to their adaptation for pollination by large bees.

Rimpau confirms Darwin, H. Miiller, and Ogle as to the self-fertilisation of our cultivated peas. Nevertheless, as is well known, marked varieties have been obtained by artificial crossing by Gartner, Knight, Laxton, and others, especially in this country.

At the same time experiments show that while it is very easy to obtain artificial hybrids of such plants, and there is no fear of natural inter-crossing, the forms are remarkably unstable as yet. Similarly unsatisfactory results were obtained with beet. As experiments are still going on, however, we may expect to hear more about these and other results.