THIS interesting plant reminds one of the handsome Japan primrose (Primula Japonica), by the manner in which the flowers are produced in a series of whorls; but the snowy primrose (P. nivalis) has the like habit, and some others indicate that a very slight change of conditions would induce them to present their flowers in a spiral arrangement, instead of a simple umbel. The Abyssinian primrose was first received in this country in the year 1825, under the name of P. involucrata, and was first figured in the Botanical Magazine in the year 1828, under t. 2842. In its original form it was a somewhat poor plant, with small flowers borne on long pedicels amidst a profusion of floral bracts and with conspicuous green calyces. Its native country was the Arabian province of Yemen, on the margins of rivulets on Kurma, a calcareous mountain in north latitude fourteen and a half degrees, that is, towards the southern extremity of Arabia Felix. A much improved form--considered from the floricultural point of view--was introduced by Messrs. Veitch and Son, of Chelsea, in the year 1872; this first flowered on the rockery at Kew in the year 1873, and was figured by Sir J. D. Hooker in the work cited above, under t. 6042. This later introduction is called Primula verticillata, var. sinensis. It is of robust habit, producing a whorl of oblong leaves, from the centre of which springs a stout flower-stem, bearing one, two, or three distinct whorls of flowers, which are larger, more richly coloured, and on shorter pedicels, with inconspicuous calyces, and therefore distinct from those of the earlier form, and considerably handsomer.

Collections of primulas are in request for rockeries, and although a few of the sorts need special and peculiar treatment, a considerable proportion of the most useful species readily conform to one simple system of cultivation. The vigorous-growing kinds require a deep sandy soil, always moist, and some amount of shade from the midday sun in the heat of summer. There is no primrose known to our gardens that can with impunity endure drought as a sempervivum or sedum can; all primroses suffer if much roasted by sunshine, and a shallow, poor soil will but rarely afford any of them a suitable root-hold. On the other hand, most of the diminutive species bear full exposure without harm, provided their roots have the advantage of a deep, moist bed. It is advisable, when collections are planted on a rockery, to associate them in groups as nearly as possible, so as to subject them to uniform treatment, and thus insure regular attention. When dotted about in places distant from each other, a few may be forgotten at times when extra attention is required. During dry hot weather water should be freely bestowed upon them, and this is more effectually accomplished when they are planted in groups than when they are distributed over a considerable space as isolated plants.

Amongst the more desirable of primulas for a rockery may be named P. auricula in its original wild form; P. capitata, in the way of P. denticulata; P. farinosa, P. integrifolia, P. latifolia, P. marginata, P. purpurea, P. rosea, P. villosa, and P. verticillata in its varietal form of sinensis.

The finest of the hardy border primulas are P. Japonica, a truly grand plant adapted for planting in masses; P. Sieboldi, which supersedes the old and much favoured P. cortusoides, and finally the varieties of our own native P. vulgaris, which are much less known in gardens than they deserve to be. All these primulas which we select for the border are suitable also for rockeries, but they are not, in a strict view of the case, rockery plants; and, moreover, to enjoy them thoroughly, they are needed in quantity, repeated and repeated in all the variety possible, from the lovely double white and crimson primroses to the rich gold-laced polyanthuses.

In the routine cultivation of primulas, the raising of stock from seed is a matter of considerable importance. All the kinds may be multiplied by division, and in the case of double varieties that do not produce seed, this is the only course of procedure possible. But division should never be resorted to if seed can be obtained, for large specimens are always to be desired, and a vigorous progeny may be best secured by resorting to seeds.