CALCEOLARIA. A BUSHY GREENHOUSE VARIETY, WITH YELLOW FLOWERS BLOTCHED WITH PURPLE.

The flowers in this genus are constructed so as to favour or almost ensure cross-fertilisation (3/6. Hildebrand as quoted by H. Muller ‘Die Befruchtung der Blumen’ 1873 page 277.); and Mr. Anderson remarks that extreme care is necessary to exclude insects in order to preserve any kind true. (3/7. ‘Gardeners’ Chronicle’ 1853 page 534.) He adds the interesting statement, that when the corolla is cut quite away, insects, as far as he has seen, never discover or visit the flowers. This plant is, however, self-fertile if insects are excluded. So few experiments were made by me, that they are hardly worth giving. Crossed and self-fertilised seeds were sown on opposite sides of a pot, and after a time the crossed seedlings slightly exceeded the self-fertilised in height. When a little further grown, the longest leaves on the former were very nearly 3 inches in length, whilst those on the self-fertilised plants were only 2 inches. Owing to an accident, and to the pot being too small, only one plant on each side grew up and flowered; the crossed plant was 19 1/2 inches in height, and the self-fertilised one 15 inches; or as 100 to 77.

Linaria vulgaris.

It has been mentioned in the introductory chapter that two large beds of this plant were raised by me many years ago from crossed and self-fertilised seeds, and that there was a conspicuous difference in height and general appearance between the two lots. The trial was afterwards repeated with more care; but as this was one of the first plants experimented on, my usual method was not followed. Seeds were taken from wild plants growing in this neighbourhood and sown in poor soil in my garden. Five plants were covered with a net, the others being left exposed to the bees, which incessantly visit the flowers of this species, and which, according to H. Muller, are the exclusive fertilisers. This excellent observer remarks that, as the stigma lies between the anthers and is mature at the same time with them, self-fertilisation is possible.(3/8.‘Die Befruchtung’ etc.page 279.) But so few seeds are produced by protected plants, that the pollen and stigma of the same flower seem to have little power of mutual interaction. The exposed plants bore numerous capsules forming solid spikes.

Five of these capsules were examined and appeared to contain an equal number of seeds; and these being counted in one capsule, were found to be 166. The five protected plants produced altogether only twenty-five capsules, of which five were much finer than all the others, and these contained an average of 23.6 seeds, with a maximum in one capsule of fifty-five. So that the number of seeds in the capsules on the exposed plants to the average number in the finest capsules on the protected plants was as 100 to 14.

Some of the spontaneously self-fertilised seeds from under the net, and some seeds from the uncovered plants naturally fertilised and almost certainly intercrossed by the bees, were sown separately in two large pots of the same size; so that the two lots of seedlings were not subjected to any mutual competition. Three of the crossed plants when in full flower were measured, but no care was taken to select the tallest plants; their heights were 7 4/8, 7 2/8, and 6 4/8 inches; averaging 7.08 in height. The three tallest of all the self-fertilised plants were then carefully selected, and their heights were 6 3/8, 5 5/8, and 5 2/8, averaging 5.75 in height. So that the naturally crossed plants were to the spontaneously self-fertilised plants in height, at least as much as 100 to 81.

Verbascum thapsus.

The flowers of this plant are frequented by various insects, chiefly by bees, for the sake of the pollen. Hermann Muller, however, has shown (‘Die Befruchtung’ etc. page 277) that V. nigrum secretes minute drops of nectar. The arrangement of the reproductive organs, though not at all complex, favours cross-fertilisation; and even distinct species are often crossed, for a greater number of naturally produced hybrids have been observed in this genus than in almost any other. (3/9. I have given a striking case of a large number of such hybrids between Verbascum thapsus and lychnitis found growing wild: ‘Journal of Linnean Society Botany’ volume 10 page 451.) Nevertheless the present species is perfectly self-fertile, if insects are excluded; for a plant protected by a net was as thickly loaded with fine capsules as the surrounding uncovered plants. Verbascum lychnitis is rather less self-fertile, for some protected plants did not yield quite so many capsules as the adjoining uncovered plants.

Plants of Verbascum thapsus had been raised for a distinct purpose from self-fertilised seeds; and some flowers on these plants were again self-fertilised, yielding seed of the second self-fertilised generation; and other flowers were crossed with pollen from a distinct plant. The seeds thus produced were sown on the opposite sides of four large pots. They germinated, however, so irregularly (the crossed seedlings generally coming up first) that I was able to save only six pairs of equal age. These when in full flower were measured, as in Table 3/25.

TABLE 3/25. Verbascum thapsus.

Heights of Plants measured in inches.

Column 1: Number (Name) of Pot.

Column 2: Crossed Plants.

Column 3: Self-fertilised Plants of the Second Generation.

Pot 1 : 76 : 53 4/8.

Pot 2 : 54 : 66.

Pot 3 : 62 : 75. Pot 3 : 60 5/8 : 30 4/8.

Pot 4 : 73 : 62. Pot 4 : 66 4/8 : 52.

Total : 392.13 : 339.00.

We here see that two of the self-fertilised plants exceed in height their crossed opponents. Nevertheless the average height of the six crossed plants is 65.34 inches, and that of the six self-fertilised plants 56.5 inches; or as 100 to 86.

Vandellia nummularifolia.

Seeds were sent to me by Mr. J. Scott from Calcutta of this small Indian weed, which bears perfect and cleistogene flowers.(3/10. The convenient term of CLEISTOGENE was proposed by Kuhn in an article on the present genus in ‘Bot. Zeitung’ 1867 page 65.) The latter are extremely small, imperfectly developed, and never expand, yet yield plenty of seeds. The perfect and open flowers are also small, of a white colour with purple marks; they generally produce seed, although the contrary has been asserted; and they do so even if protected from insects. They have a rather complicated structure, and appear to be adapted for cross-fertilisation, but were not carefully examined by me. They are not easy to fertilise artificially, and it is possible that some of the flowers which I thought that I had succeeded in crossing were afterwards spontaneously self-fertilised under the net. Sixteen capsules from the crossed perfect flowers contained on an average ninety-three seeds (with a maximum in one capsule of 137), and thirteen capsules from the self-fertilised perfect flowers contained sixty-two seeds (with a maximum in one capsule of 135); or as 100 to 67.

But I suspect that this considerable excess was accidental, as on one occasion nine crossed capsules were compared with seven self-fertilised capsules (both included in the above number), and they contained almost exactly the same average number of seed. I may add that fifteen capsules from self-fertilised cleistogene flowers contained on an average sixty-four seeds, with a maximum in one of eighty-seven.

Crossed and self-fertilised seeds from the perfect flowers, and other seeds from the self-fertilised cleistogene flowers, were sown in five pots, each divided superficially into three compartments. The seedlings were thinned at an early age, so that twenty plants were left in each of the three divisions. The crossed plants when in full flower averaged 4.3 inches, and the self-fertilised plants from the perfect flowers 4.27 inches in height; or as 100 to 99. The self-fertilised plants from the cleistogene flowers averaged 4.06 inches in height; so that the crossed were in height to these latter plants as 100 to 94.

I determined to compare again the growth of plants raised from crossed and self-fertilised perfect flowers, and obtained two fresh lots of seeds. These were sown on opposite sides of five pots, but they were not sufficiently thinned, so that they grew rather crowded. When fully grown, all those above 2 inches in height were selected, all below this standard being rejected; the former consisted of forty-seven crossed and forty-one self-fertilised plants; thus a greater number of the crossed than of the self-fertilised plants grew to a height of above 2 inches. Of the crossed plants, the twenty-four tallest were on an average 3.6 inches in height; whilst the twenty-four tallest self-fertilised plants were 3.38 inches in average height; or as 100 to 94. All these plants were then cut down close to the ground, and the forty-seven crossed plants weighed 1090.3 grains, and the forty-one self-fertilised plants weighed 887.4 grains. Therefore an equal number of crossed and self-fertilised would have been to each other in weight as 100 to 97.

From these several facts we may conclude that the crossed plants had some real, though very slight, advantage in height and weight over the self-fertilised plants, when grown in competition with one another.

The crossed plants were, however, inferior in fertility to the self-fertilised. Six of the finest plants were selected out of the forty-seven crossed plants, and six out of the forty-one self-fertilised plants; and the former produced 598 capsules, whilst the latter or self-fertilised plants produced 752 capsules. All these capsules were the product of cleistogene flowers, for the plants did not bear during the whole of this season any perfect flowers. The seeds were counted in ten cleistogene capsules produced by crossed plants, and their average number was 46.4 per capsule; whilst the number in ten cleistogene capsules produced by the self-fertilised plants was 49.4; or as 100 to 106.

3. GESNERIACEAE.—Gesneria pendulina.

In Gesneria the several parts of the flower are arranged on nearly the same plan as in Digitalis, and most or all of the species are dichogamous. (3/11. Dr. Ogle ‘Popular Science Review’ January 1870 page 51.) Plants were raised from seed sent me by Fritz Muller from South Brazil. Seven flowers were crossed with pollen from a distinct plant, and produced seven capsules containing by weight 3.01 grains of seeds. Seven flowers on the same plants were fertilised with their own pollen, and their seven capsules contained exactly the same weight of seeds. Germinating seeds were planted on opposite sides of four pots, and when fully grown measured to the tips of their leaves.

TABLE 3/26. Gesneria pendulina.

Heights of Plants measured in inches.

Column 1: Number (Name) of Pot.

Column 2: Crossed Plants.

Column 3: Self-fertilised Plants.

Pot 1 : 42 2/8 : 39. Pot 1 : 24 4/8 : 27 3/8.

Pot 2 : 33 : 30 6/8. Pot 2 : 27 : 19 2/8.

Pot 3 : 33 4/8 : 31 7/8. Pot 3 : 29 4/8 : 28 6/8.

Pot 4 : 30 6/8 : 29 6/8. Pot 4 : 36 : 26 3/8.

Total : 256.50 : 233.13.

The average height of the eight crossed plants is 32.06 inches, and that of the eight self-fertilised plants 29.14; or as 100 to 90.

4. LABIATAE.—Salvia coccinea. (3/12. The admirable mechanical adaptations in this genus for favouring or ensuring cross-fertilisation, have been fully described by Sprengel, Hildebrand, Delpino, H. Muller, Ogle, and others, in their several works.)

This species, unlike most of the others in the same genus, yields a good many seeds when insects are excluded. I gathered ninety-eight capsules produced by flowers spontaneously self-fertilised under a net, and they contained on an average 1.45 seeds, whilst flowers artificially fertilised with their own pollen, in which case the stigma will have received plenty of pollen, yielded on an average 3.3 seeds, or more than twice as many. Twenty flowers were crossed with pollen from a distinct plant, and twenty-six were self-fertilised. There was no great difference in the proportional number of flowers which produced capsules by these two processes, or in the number of the contained seeds, or in the weight of an equal number of seeds.

Seeds of both kinds were sown rather thickly on opposite sides of three pots. When the seedlings were about 3 inches in height, the crossed showed a slight advantage over the self-fertilised. When two-thirds grown, the two tallest plants on each side of each pot were measured; the crossed averaged 16.37 inches, and the self-fertilised 11.75 in height; or as 100 to 71. When the plants were fully grown and had done flowering, the two tallest plants on each side were again measured, with the results shown in Table 3/27.