.... the lover of flowers who visits the island will find that the details hidden in the long-distance view from the mainland are many and exciting. A close examination by the knowledgeable amateur will enable him to make an interesting list of the flowers, interesting not because of the great number of rare species but rather because there are few species, some of which are often individually very numerous and may form pure communities. The more serious student of plant ecology will instantly recognise the characteristic grouping of species into associations or societies peculiar to certain types of country; and the general naturalist finds that certain species of mammals, birds, insects and other creatures frequent these territories dominated by one or two typical plants. This linking up of animal and plant with terrain can be most conveniently studied on a small sub-Atlantic island, where the limitations of isolation and area seem to make possible an intensive investigation of its restricted ecology. Skomer exhibits a very considerable diversity of plant habitats, due to several factors: the topography of its high plateau with steep cliff walls, the imperviousness and acidity of much of the volcanic rock, the varying proximity of the sea, the climate with its frequent fierce winds, the alkaline and nitrogenous droppings of the seabirds, the effects of overgrazing by rabbits. We began our study by making a card-indexed list of the plants from specimens collected. But we found it difficult to place species into strict ecological communities, owing to the lack, over considerable areas, of a dominant plant. There were, however, obviously two general categories: inland and maritime. As a guide the following eleven main ecological divisions or habitats may be cited: inland-sheltered well- drained land with fairly deep soil (bracken areas), inland pasture, dry turf, heath, rock outcrops, ponds and streams, bog, and human habitation (buildings, walls, garden); maritime - exposed cliff-tops of the south and west (favoured by thrift), less exposed cliff-tops of the north and east (favoured by sea-campion and scurvy-grass), and steep sea-cliffs. We fully recognise the inadequate nature of such a classification, and do not propose to labour it further in this chapter; since the reports we received from our botanists showed clearly that they too had failed to set up more than a very elastic system. We shall confine ourselves therefore to a general description of the plant ecology.

The plant communities were closely watched from the end of March onwards to October, and their seasonal changes followed closely upon those recorded in parallel studies at Skokholm. But Skomer has, from its greater size, its configuration, and greater diversity of soils, a richer flora than that of Skokholm.

The earliest flowers noted were those of the lesser celandine (Ranunculus Ficaria), ground-ivy (Glechoma hederacea), primrose (Primuta vulgaris), and scurvy-grass (Cochlearia officinalis); all, except the last, forming part of the bracken (Pterilium aquilinum) society though they are not confined to it.

There are two footnotes. Specimens of the very large variety of the buck's-horn plantain (Plantago Coronopus) with almost entire leaves attracted the attention of several of the amateur botanists who visited the Centre in 1946, including Mr. Stephen Potter and Dr. Julian Huxley. Specimens and seeds were sent to Kew where they were grown in the experimental ground there. Dr. W. B. Turrill, Keeper of the Herbarium at Kew, reported on June 3rd, 1949, that these specimens retained their characters under cultivation and the essential ones reappeared in plants grown from seed to at least the third generation. He considers that the plant is var. maritima Gren. et Godr.