Abstract: the climatic factors, though for purposes of separation and classification of systems it is a legitimate procedure. In fact the climatic complex has more effect on the organisms and on the soil of an ecosystem than these have on the climatic complex, but the reciprocal action is not wholly absent. Climate acts on the ecosystem rather like an acid or an alkaline " buff er " on a chemical soil complex. Next comes the soil complex which is created and developed partly by the subjacent rock, partly by climate, and partly by the biome. Relative maturity of the soil complex, conditioned alike by climate, by subsoil, by physiography and by the vegetation, may be reached at a different time from that at which the vegetation attains its climax. Owing to the much greater local variation of subsoil and physiography than of climate, and to the fact that some of the 4The mental isolates we make are by no means all coincident with physical systems, though many of them are, and the ecosystems among them. This content downloaded from 220.127.116.11 on Thu, 11 Aug 2016 05:29:06 UTC All use subject to http://about.jstor.org/terms July, 1935 VEGETATIONAL CONCEPTS AND TERMS 301 existing variants prevent the climatic factors from playing the full part of which they are capable, the developing soil complex, jointly with climate, may determine variants of the biome. Phillips' contention that soil never does this is too flatly contrary to the experience of too many ecologists to be admitted. Hence we must recognise ecosystems differentiated by soil complexes, subordinate to those primarily determined by climate, but none the less real. Finally comes the organism-complex or biome, in which the vegetation is of primary importance, except in certain cases, for example many marine ecosystems. The primary importance of vegetation is what we should expect when we consider the complete dependence, direct or indirect, of animals upon plants. This fact cannot be altered or gainsaid, however loud the trumpets of the " biotic community " are blown. This is not to say that animals may not have important effects on the vegetation and thus on the whole organismcomplex. They may even alter the primary structure of the climax vegetation, but usually they certainly do not. By all means let animal and plant ecologists study the composition, structure, and behaviour of the biome together. Until they have done so we shall not be in possession of the facts which alone will enable us to get a true and complete picture of the life of the biome, for both animals and plants are components. But is it really necessary to formulate the unnatural conception of biotic comimiunity to get such co-operative work carried out? I think not. What we have to deal with is a system, of which plants and animals are components, though not the only components. The biome is determined by climate and soil and in its turn reacts, sometimes and to some extent on climate, always on soil. Clements' "prisere " ('16) is the gradual development of an ecosystem as we may see it taking place before us to-day. The gradual attainment of more complete dynamic equilibrium (which Phillips quite rightly stresses) is the fundamental characteristic of this development. It is a particular case of the universal process of the evolution of systems in dynamic equilibrium. The equilibrium attained is however never quite perfect: its degree of perfection is measured by its stability. The atoms of the chemical elements of low atomic number are examples of exceptionally stable systems-they have existed for many millions of millennia: those of the radio-active elements are decidedly less stable. But the order of stability of all the chemical elements is of course immensely higher than that of an ecosystem, which consists of components that are themselves more or less unstable-climate, soil and organisms. Relatively to the more stable systems the ecosystems are extremely vulnerable, both on account of their own unstable components and because they are very liable to invasion by the components of other systems. Nevertheless some of the fully developed systems-the " climaxes "-have actually maintained themselves for thousands of years. In others there are elements whose slow change will ultimately bring about the disintegration of the system. This content downloaded from 18.104.22.168 on Thu, 11 Aug 2016 05:29:06 UTC All use subject to http://about.jstor.org/terms 302 A. G. TANSLEY Ecology, Vol. 16, No. 3 This relative instability of the ecosystem, due to the imperfections of its equilibrium, is of all degrees of magnitude, and our means of appreciating and measuring it are still very rudimentary. Many systems (represented by vegetation climaxes) which appear to be stable during the period for which they have been under accurate observation may in reality have been slowly changing all the time, because the changes effected have been too slight to be noted by observers. Many ecologists hold that all vegetation is always changing. It may be so: we do not know enough either to affirm or to deny so sweeping a statement. But there may clearly be minor changes within a system which do not bring about the destruction of the system as such. Owing to the position of the climate-complexes as primary determinants of the major ecosystems, a marked change of climate must bring about destruction of the ecosystem of any given geographical region, and its replacement by another. This is the clisere of Clements ('16). If a continental icesheet slowly and continuously advances or recedes over a considerable period of time all the zoned climaxes which are subjected to the decreasing or increasing temperature will, according to Clements' conception, move across the continent " as if they were strung on a string," much as the plant communities zoned round a lake will move towards its centre as the lake fills up. If on the other hand a whole continent desiccates or freezes many of the ecosystems which formerly occupied it will be destroyed altogether. Thus whereas the prisere is the development of a single ecosystem in situ, the clisere involves their destruction or bodily shifting. When we consider long periods of geological time we must naturally also take into account the progressive evolution and rise to dominance of new types of organism and the decline and disappearance of older types. From the earlier Palaeozoic, where we get the first glimpses of the constitution of the organic world, through the later Palaeozoic where we can form some fairly comprehensive picture of what it was like, through the Mesozoic where we witness the decline and dying out of the dominant Palaeozoic groups and the rise to prominence of others, the Tertiary with its overwhelming dominance of Angiosperms, and finally the Pleistocene ice-age with its disastrous results for much of the life of the northern hemisphere, the shifting panorama of the organic world presents us with an infinitely complex history of the formation and destruction of ecosystems, conditioned not only by radical changes of land surface and climate but by the supply of constantly fresh organic components. We can never hope to achieve more than a fragmentary view of this history, though doubtless our knowledge will be very greatly extended in the future, as it has been already notably extended during the last 30 years. In detail the initiation and development of the ecosystems in past times must have been governed by the same principles that we can recognize to-day. But we gain nothing by trying to envisage in the same concepts such very different processes as are involved in the shifting or destruction of ecosystems on the one hand and the development of individual systems on the This content downloaded from 22.214.171.124 on Thu, 11 Aug 2016 05:29:06 UTC All use subject to http://about.jstor.org/terms July, 1935 VEGETATIONAL CONCEPTS AND TERMS 303 other. It is true, as Cooper insists ('26), that the changes of vegetation on the earth's surface form a continuous story: they form in fact only a part of the story of the changes of the surface of this planet. But to analyse them effectively we must split up the story and try to focus its phases according to the various kinds of process involved.