The work of George W. Oliver in the greenhouses of the Department of Agriculture at Washington has thrown much light on the requirements of young mangosteen plants, and on the best methods of propagation. The following extracts are taken from his report in Bulletin 202 of the Bureau of Plant Industry, "The Seedling-inarch and Nurse-plant Methods of Propagation."

"Few plants show the results of inattention on the part of the cultivator more plainly than the mangosteen. When once a plant becomes in the least sickly, there is little likelihood of its recovery on its own roots. The mangosteen does not take kindly to heavy soils; it prefers a well-drained soil containing a large proportion of decayed vegetable matter. When seedlings are removed from flats and put in pots some will die without apparent cause. An over-supply of water causing the soil to become in the least sour is certain to induce sickness much more quickly in the mangosteen than in other species of the genus. Therefore, great care is necessary in handling the plants, especially in the early stages of the seedlings.

"Unfortunately the mangosteen is not a strong-rooting plant, especially during the first year or two after germination. This peculiarity renders it particularly sensitive to dry weather and may account in part for the many failures to grow it successfully. Nearly all the other species of the genus have strong and abundant roots, even in the seedling stages. It therefore seems likely that the mangosteen will thrive better and under more widely varying soil and atmospheric conditions if the young plant is inarched to some species of the genus which has a good root system.

' The genus Garcinia is a large one, the Index Kewensis listing 228 species. Of these about 20 have been tried in the inarching experiments; and while the mangosteen unites with all of them, only a few can be recommended as promising stock-plants. Two other genera of the same family, Calophyllum and Platonia, have been tried. Two species of Calophyllum, C. calaba and C. inophyllum, are not satisfactory because the union between these and the mangosteen is imperfect. This is partly because the stems of the Calophyllums are softer than those of the seedling mangosteen and partly because the growth made by the former as they become older is much more rapid. Platonia insignis (see below), on the other hand, so far as the experimental work has progressed, is a very promising stock from one to three years after germination, and if it will grow under the conditions suitable for the mangosteen, it may turn out to be the best stock of all those tried. The most promising species of Garcinia for use as stock-plants for the mangosteen are G. tinctoria, G. morella, and G. Livingstonei, in the order named, the last a native of Portuguese East Africa. The two first named are from the Malay Peninsula."

Recent experiments have shown that G. xanthochymus is also promising. It is vigorous in growth, and adapted to many types of soil. Inarching the mangosteen is a simple process, essentially the same as inarching the mango. Oliver says further:

"None of the species of Garcinia used as stocks are difficult to raise from seeds, provided they are fresh. They are easiest to germinate when sown in soil composed largely of partially decomposed leaves mixed with a little loam and rough-grained sand. They should be potted as soon as the first leaves are well developed. All the Garcinias with the exception of G. Mangostana have magnificent root systems and they thrive under ordinary treatment in so far as soil watering and a considerable range of temperature are concerned.

"It is an important point to have the stock plants in an active stage of growth when the union is in progress, though the seedling mangosteens may be inarched while apparently dormant. Although the unions when both stock and cion are in a resting stage are fairly satisfactory, the difference in growth is easily observable when the stock plants are in active growth. To secure this active growth the stocks should be allowed to become dormant; then, when they are given larger pots, good drainage, and soil composed of rotted leaves, at least one-half, and the rest fibrous loam containing a little rough-grained sand, together with some rough charcoal and crushed bone, they will under high temperature respond with vigorous growth. When inarched in this condition the union is always satisfactory.

" All plants used as stocks have been from one to three years old. Within that period the age of the stocks seems to make little difference, especially when used as nurse stocks. Mangosteen seedlings seven months old united on nurse-stocks of three-year-old Garcinia tinctoria made very fine unions, and within six months after the union some of the mangosteen stems were almost as thick as those of the stocks."

P. J. Wester states that the mangosteen can be budded, and says : "Use mature, green and smooth, nonpetioled budwood; cut the buds an inch and a half long; insert the buds in the stock at a point of the same appearance as the cion or at most where it is streaked with gray."

These methods of propagating the mangosteen are of recent inception but they promise to be of immense value in extending the area in which the tree can be grown, as well as in permitting the establishment of superior varieties, which is not possible when seed-propagation is the only means used. Fairchild writes: "When one considers that so far no selection of varieties of the mangosteen has been made, notwithstanding the fact that practically seedless fruits are of frequent occurrence, and further that the tree belongs to a large genus of fruit-bearing trees, at least fifteen of which are known to bear edible fruits, some of them as large as small melons, and that these are scattered in Australia, the Malay region, South China, Africa, Brazil, and Central America, it becomes evident that in the development and breeding of the mangosteen and in the discovery of a suitable stock for it, there lies a most promising field for horticultural research."