Nature Vol. 295 21 January 1982 189 in an animal tumour system selective aggregation and is thought to be a direct that the early part of the post-glacial in inhibitors of Tx, which promotes platelet effect on the rate of proliferation of the north-west (pre 7,000 BP) was aggregation, and the administration of tumour cells by increasing cyclic AMP. characterized by a vigorous circulation of

PGl2 (or of substances which facilitate The meeting was successful in bringing air masses which would have brought dry, vascular PGI2 synthesis), which prevents together cancer research workers and many anticyclonic conditions to the area both in the formation of platelet emboli. Both of the leaders in the area of prostaglandins. winter and summer. It may well be that procedures reduced the incidence of lung It was encouraging to both groups that many southern continental species metastasis and in addition slowed the manipulation of the AA cascade may have were able to migrate into what are now growth of the tumour growing at the site of a role in both prevention and treatment of oceanic regions in the north and west injection. This latter effect is unlikely to cancer although the key experiments during those times. This would have involve interference with platelet remain to be performed. D brought into contact a variety of continental species of somewhat southerly and northerly affinities; it is interesting to note that all five of the species under discussion here are currently found together in central Poland, which may give A new British flower some indication of the climatic conditions during the period of mixing. Rix and from Peter D. Moore W oods1 note that the Stanner Rocks specimens of G. bohemica are mor­ BoTANISTS in the tropics will probably having a number of open habitat species of phologically most akin to the western throw up their hands in horror but the fact generally disjunct distribution were often French populations on the continent, remains that the discovery of a new species associated with disturbed or unstable sites which may suggest a west European origin of in the British Isles (and in which the forest cover had remained for the British population of this species. probably a native one at that) is news. Rix incomplete throughout post-glacial time. The subsequent history of the mid Wales and W oods1 have now officially recorded The now isolated in such sites may area has involved forest spread and the finding of bohemica at Stanner represent relicts of the vegetation of late­ increasing oceanicity of climate, both of Rocks, near Kington, mid Wales. Its incon­ glacial or early post-glacial times, which which will have operated against the spicuous leaves and flower undoubtedly have been extinguished over the bulk of survival of all these species; but the rock contributed to its success in eluding their former range by forest expansion and outcrops, such as Stanner, have evidently detection until recently and to its initial consequent shading. In the case of Stanner acted as refuges. Such outcrops would not misidentification. But the find also has only have provided steep, unstable slopes some interesting biogeographical implica­ keeping the woodland and canopy open tions, for G. bohemica is a species of and discontinuous, but may also have continental distribution, reaching its created microclimatic conditions northwestern European limits in central appropriate for the survival of continental Germany and the Seine Valley in France, species of plant. G. bohemica on Stanner and extending into central, eastern and Rocks is currently restricted to shallow southern parts of the European mainland. soils on southern and eastern aspects. Stanner Rocks is a site far-removed from These locations would be well insolated in the remainder of its range. summer, providing a 'micro-continental' One must consider the possibility that climate, rather analogous to some of the the plant has arrived at this location by present-day East Anglian breckland chance dispersal but, as Carlquist2 has habitats of S. perennis and V. spicata. The recently emphasized, it is very difficult to importance of aspect for survival of a study freak phenomena of this kind. G. species at the edge of its climatic range is bohemica reproduces by seed (though at well documented for such plants as Cirsium Stanner Rocks only one per cent of the acaule8 which reaches its northern British population flowers and no capsules have limit in Derbyshire but survives there been observed) and bybulbils. Unlike some mainly on southerly aspects. Even under lake and seashore sites, a rock outcrop is an these conditions fruiting is poor and unlikely spot for the concentration of a experiments involving daily spraying of the bird migration route, making bird developing fruit with water, thus lowering transport an unlikely explanation of its overall temperatures, reduced success even presence. Gagea bohemica further. G. bohemica fails to fruit at It is important to consider that Stanner Rocks, such an explanation is supported by Stanner, but persists by vegetative Rocks boasts an unusual assemblage of the sub-fossil record of certain of the plants propagation. rare plants currently exhibiting disjunct concerned. S. perennis pollen has been The record of G. bohemica from Stanner distributions, including Lychnis viscaria, found in late-glacial sediments of mid Rocks is thus more than just one more Veronica spicata and Scleranthus perennis. Wales 5 and a seed of L. viscaria was found species on our list - it may also provide Also, on some neighbouring outcrops in by Dickson, Dickson and Mitchell6 in late­ important evidence of the climatic and mid Wales, is Potentil/a rupestris3 • All of Devensian material from the Isle of Man. vegetational history of the British Isles. D these have continental distributions, Unfortunately, no sub-fossil records are though L. viscaria and S. perennis extend available for V. spicata, P. rupestris or G. I. Rix & Woods Watsonia 13,265 (1981). further north than the others. bohemica in Britain, but their current, 2. Carlquist Am. Scient. 69,509 (1981). All these species are plants of low growth 3. Sinker in Welsh Wildlife in Trust (ed. Lacey, W.S.) 129 rather southerly, continental distributions (North Wales Naturalists Trust, 1970). habit and demand light. Pigott and do not make them likely candidates for 4. Pigott & Walters J. Ecol. 42, 95 (1954). Walters4 proposed that plant assemblages late-glacial relicts. 5. Moore New Phyto/. 69, 363 (1970). 6. Dickson, Dickson & Mitchell Phil. Trans. R. Soc. B258, Lamb7 has constructed global maps of Peter D. Moore is Senior Lecturer in the 31 (1970). probable atmospheric pressure systems 7. Lamb in Climate: Present, Past and Future Vol. 2, 382 Department of Plant Sciences, University of (Methuen, London, 1977). London King's College, 68 Half Moon Lane, during winter and summer at several stages 8. Pigott in Flora of a Changing Britain (ed. Perring, F.) 32 London SE24 9JF. of post-glacial history, and he concludes (Classey, Hampton, Middlesex, 1970).

0028-0836/82/030189-01$01.00 © 1982 Macmillan Journals Ltd