III. Notice of Plants at Dr Shapter's Garden at Cobham, near Exeter. By Professor BALFOUR. (Plate I.)

(Read 5th December 1876.)

When paying a visit in September 1876 to my old friend Dr Shapter at Exeter, I had an opportunity of seeing his garden arrangements, and of observing the success with which he cultivated many valuable plants which thrive in the climate of Devonshire.

Dr Shapter commenced his botanical studies in the University of Edinburgh, under Dr Graham, in whose excursions he and I joined. He visited many stations in the neighbourhood of Edinburgh, and also examined the Flora of our alpine districts. He has continued to take an interest in botany, and while engaged in laborious practice he has found opportunities of cultivating his taste for Natural History.

His garden is at Cobham, where he has a cottage, to which he repairs in the summer evenings, returning to his practice in Exeter in the morning-enjoying thus both a town and a country life. The garden is on the north slope of the high ground to the north-east of Exeter, at an elevation of 350 feet above the level of the sea, and of which there is an uninterrupted view ten miles distant. The garden is freely exposed to the east, south, and west, and is slightly protected on the north by the corner of the hill. There are in the garden a great number of fine Coniferous plants, specimens of Palms and Bamboos, New Zealand Flax, Arundo Donax, Aralias, Agaves, and Cordylines. What especially attracted my notice was a fine specimen of Cordyline (Dracaena) australis. Dr Shapterinformed me that the plant was only twelve years old, and had been raised by his gardener, John Portbury, an active, intelligent man of the old school. The stem of the plant is 10 feet 6 inches in height and 2 feet 1 inch in circumference. The cluster of leaves at the top has a circumference of 21 feet, the leafy part of the stem being 6 feet long. The leaves average about 4 feet 2 inches in length. The leaves have been annually tied together from November to the first week of March, so prevent rain, snow, or frost penetrating to the central

as to

part of the stem. The flowering commenced in May 1876, and the full development of the floral clusters took place about the 30th of June. The stalk bearing the cluster was about 4 feet long, and 7 inches in circumference, a foot above where it emerged from the stem. From the base of the stem to the summit of the flowering panicle the height is 17 feet. The flowers are white, and very fragrant. The sheathing part of the leaf at the base is on an average 4 inches in breadth, it then narrows rapidly, and at six inches from the base it is only 1 inch. It gradually expands to 2 inches about two feet from the stem, and then tapers to a point. The plant, of which I showed a photograph at last meeting, is a very remarkable one as regards its growth and luxuriance, and is a good indication of the nature of the climate and the exposure. (Plate I.)

In the same garden there are several Palms, which thrive well, such as Chamaerops humilis, and C. excelsa.

Another plant of great interest, as regards its growth, is Cupressus macrocarpa, a photograph of which I showed at last meeting. There are several fine specimens of this Cupressus in the garden, but that which calls for special attention was planted not far from the Cordyline. It was put into the garden about 20 years ago, a moderate-sized shrub, and now it is upwards of 51 feet high. (Plate I.)

Of the Bamboos in the garden, Dr Shapter pointed out Bambusa Metake, which he said was always in leaf, B. falcata, the canes of which attain a height of 17 feet-in flower only, not having developed this summer a single leaf. He pointed out thriving plants of Aralia japonica and Phormium tenax; and he stated that the Arundo Donax sprouts up every year to the height of 12 or 13 feet.

I was very much struck with the growth of the plants in this favoured spot. The nursery grounds near Exeter are also objects of great interest, and well deserve a visit from Horticulturists and Arboriculturists. I have already brought under the notice of the Society the dimensions of some of the trees cultivated in Pince's Nurseries at Exeter.

IV. Experiments with Turnip Seeds. By A. STEPHEN WILSON. (Plate II.)

(Read 8th February 1877.)

The present paper contains the results of experiments proposed to be made by the writer, in a communication laid before the Botanical Society in January 1876. That communication detailed an experiment in which large and small seeds of Swedish turnip (Brassica campestris rutabaga) were grown in water, on opposite sides of the same plate, and in which the larger seeds produced the larger plants; suggesting the probability that a part of the difference in the sizes of contiguous bulbs in ordinary field culture is due to difference in the size of the seeds from which these bulbs have been produced. From certain experiments, Darwin has come to the general conclusion ("Cross and Self-Fertilisation of Plants," p. 353), that "it cannot be doubted that heavy and fine seeds tend to yield the finest plants."

A full discussion of the anatomy and physiology of the turnip, from the embryo to the flower, although very desirable in connection with these experiments, must for the present be deferred.

The common turnip of many varieties (Brassica rapa) as well as the Swedish turnip, are usually stated to be biennial plants. But a precise and unqualified statement of this kind is likely to mislead, inducing the conception that bienniality is essential to the existence of these plants. Now, whether a turnip shall be annual or biennial depends upon the circumstances under which it is grown, and perhaps also, in some cases, upon inherited tendencies. A look across the fields or gardens, almost any season, will reveal here and there a plant running to seed. If such a plant is examined, it will usually be found to have a small or defective bulb. The law therefore seems to be, that if bulbing is prevented, annuality is secured. While if the leaves and leaf-stalks which form the stems, are arrested below the insertion of the cotyledons, they will be fleshified and swelled out into a bulb, and the plant will be made biennial.

Some further experiments were made by the writer, of

the same character as those detailed in his last paper. The saucer in which the plants were growing, in water, and the dust which fell upon them, was set aside in the window. The seeds were put into the water on 8th January (1876); many of them came into flower in June at a height of 3 or 4 inches, some of them ripening one seed and some two seeds, of nearly the ordinary size (Pl. II. fig. 1); not one of these saucer plants showed the least appearance of bulbing. Bulbing, therefore, is not an essential fact of the turnip, but a fact the cause of which may be inquired into, with the purpose of controlling that cause in a direction beneficial to agriculture.

In selecting the large and small seeds for these experiments, both sets were taken from the same parcels. Probably this is a condition of importance, because equal seeds from different stocks may have unequal tendencies to bulb. It is generally believed, that seed from transplanted turnips are productive of a heavier crop than seed from untransplanted. The seeds selected were the largest and the smallest unshrivelled seeds. They were not compared as to weight; but, as in my first experiment, the large seeds bore to the small the ratio of 100 to 59, a ratio near this may be taken as descriptive of the classes of seeds below.

Experiment 1.-On the 22d May, in newly dug ground in the garden, two rows of swedes, large and small, a single seed in each hole of an inch deep, were planted at distances 8 inches apart.

Except two plants, opposite each other, all the others were either completely eaten by the fly, or so much damaged as to be put out of competition. The blanks could have been filled up, but as this would have given plants grown at different dates, and with different length of life, the blanks were allowed to remain. The two bulbs which competed were pulled on the 8th January. The net weight from the large seed was 3 lb. 15 oz.; from the small seed, 4 lb. 3 oz.

Experiment 2.-On the 24th May, in the field, 40 yards of two adjoining drills were planted with large and small Swedish seeds. Two or three seeds were put in each hole at distances intended to be 8 inches apart, but there were

The

180 holes on the small side and 170 on the large. seeds were deposited an inch deep, on drills newly manured and closed and rolled down by the machine, with the seed boxes shut. The seeds in every hole germinated, but the young plants were severely handled by the fly. By the 8th June many were destroyed; so that wherever the plants appeared to be so much damaged as to make it unlikely that they would come through, the stumps were removed, and large seeds of yellow turnips put in the blanks of the large-seed drill, and small seeds of the same yellow turnips put in the blanks of the small-seed drill, for another experiment. On the 30th of June both swedes and yellows were singled by the hand, all weeds being pulled out, and the drills being left unreduced. They were singled on the principle of leaving the best plant in each hole; but in the case of the swedes, the fly in many instances had left but a single plant of the two or three which came up. Of the large-seed swedes, the fly left 67 per cent.; of the small, 58 per cent. When the singling had been finished, it was obvious to the eye that the large seeds had the advantage. Of these there were 66 per cent. of good fair plants; while of the small seeds there were of good fair plants but 41 per cent. Hardly any of the swedes had wholly escaped the fly. The sides of the drills were hoed a second time, and the weeds on the top pulled out, but the drills were not horse-hoed, nor the earth taken away from the bulbs. From various causes, such as insects, wind, &c., a few more of the plants were destroyed after singling, but no blanks were made up.

The

The swedes were pulled on the 20th January. number of large seeds was 105; the gross weight 307 lb. 12 oz.; the net weight, 300 lb. 4 oz.; the average net weight was 2 lb. 13 oz. The heaviest bulb weighed 7 lb. 10 oz.; the lightest, 6 oz. The ten heaviest weighed 53 lb. 6 oz.; the ten lightest weighed 9 lb. 9 oz.

The number of small seeds was 108; the gross weight 267 lb. 9 oz.; the net weight 262 lb. 4 oz.; the average weight 2 lb. 6 oz. The heaviest bulb weighed 5 lb. 14 oz.; the lightest, 2 oz. The ten heaviest weighed 47 lb. 8 oz.; the ten lightest weighed 5 lb.

Experiment 2-A.-The green-top yellows, which were