Int'l Hort. Exhibition 1866, pp. 142-144


IT has been a habit with me for many years, while engaged in horticultural pursuits, to examine minutely the seeds of a plant before committing them to the earth, being convinced that much has yet to be learned, from this practice alone. Amongst other things I have eagerly watched for any symptom of change in the appearance of seeds contained in such fruits as have been fertilised by other than their own pollen. During, however, a long course of observation, only one fact of interest has presented itself bearing upon this subject, but this one is of so curious a nature as to be, I think, well worthy of the attention of physiologists.

About ten years since I took in hand the genus Matthiola, mainly for the purpose of improving it for garden purposes by cross-breeding, and of trying at the same time if any light could be thrown upon the subject of doubleness in flowers. Nothing, however, was effected in this direction worth mentioning.

Our large red-flowered biennial garden Stock, "Cocardeau" of the French, bears seeds of a light brown uniform tint, while, in the case of the purple branching "Queen Stock," a distinct race near the normal M. incana, they are of a violet black. Blossoms of the red Cocardeau having been fertilised by the pollen of the purple kind, the ripened pods were examined, and found to contain about fifty per cent. of black seeds.

The black and brown seeds, contained in one pod, were then sown in separate pots, and my satisfaction may be imagined, when I saw the young plants from the brown seed coming up with green stems, while those from the black were strongly tinged with purple. There was now no doubt of the ultimate result. The latter plants produced blossoms of a rich purple, while the former scarcely differed in habit, and not at all in colour, from the red seed-bearing parent. My observations upon the derivatives of these plants were not made with sufficient accuracy to enable me to speak on that subject, but the same curious habit of producing both colours from one seed-pod was certainly transmitted in the case of the purples through several generations.

The pre-potency, however, of the purple colouring matter was apparent, and is worthy of remark. Repeated infusions of red, or Cocardeau blood, only seemed to deepen, and enrich the purple colour, and it was not until after many generations of crossing, that a really red purple was obtained. I may add that I forwarded unopened seed vessels to Mr. Darwin, and that my results were verified by that distinguished physiologist.

Now, were the red seedlings pure emanations from the mother parent unaffected by the foreign pollen? I think not, for at the same period, I raised plants from the little annual glabrous-leaved Stock of the nurserymen, by the pollen of the large sort above mentioned. The effect was that one-half of the seedlings were glabrous, or "Wallflower-leaved," and the rest rough, like the male parent, no intermediate form occurring—that is, as far as texture of leaf was concerned; but the glabrous seedlings were no longer "miniature" Stocks, being of tall and strong habit. In this case also succeeding generations raised from rough-leaved derivatives produced a per centage of glabrous plants. The converse was not observed, and, I think, did not occur. These experiments were verified by repeated trials, followed by the same results in every case.

NOTE.—Since writing the above a fresh instance of the pre-potency of the purple colouring matter has come under my notice. Some years ago I made many crosses between red and purple sweet peas. The seedlings always came purple, and I concluded that, from imperfect manipulation, or the agency of insects, the cross fertilisation had not taken place. Last summer, however, I crossed a blossom of the beautiful new red sweet pea, called "Invincible," with the pollen of the common purple sort. All the seedlings are now (August, 1866) in blossom, and cannot be distinguished from the male or pollen parent. In this instance every precaution was taken to make an accurate experiment.