Prior to total exploitation in the mid-19th century, the principal guano islands of the Peruvian coast bore deposits of consolidated bird-droppings up to about 50 metres in thickness. The archaeological evidence assembled by Professor G. A. Kubler and the stratigraphic evidence brought together by the speaker indicate that this huge accumulation could have formed in a few thousand years. Prior to initiation of deposition on the Peruvian coast there was certainly much guano being laid down on the coast of, Chile, but farther north where the main deposits are formed to-day it is probable that the prevailing climatic regime was that of modern ‘abnormal’ wet years. In the central Pacific phosphatic guano deposits of post-glacial date and of large size, exceeding 10,000 tons, are confined to the area of low rainfall along the equator, within the zone of the South-East trade-winds and in the neighbourhood of the equatorial divergence. There is, however, reason to suppose that in such a region deposits of the size implied by the phosphate beds of Nauru, Ocean Island and some other localities in the Pacific, Christmas Island in the Indian Ocean and some West Indian Islands, could not have formed in a reasonable time. It is suggested that some huge deposits originated in periods of greatly enhanced oceanic fertility, due to a lowering of the vertical stability of the oceans when the supply of cold water was reduced during interglacials. The method used in plant-geographical analysis of a certain floristic area is the division of the plants in distributional units called ‘elements’. The term ‘element’ which was first introduced by H. Christ (1867) was subject to a large measure of conflicting interpretations. To mention only the misuse of the term ‘pontic element’ originally applied. by Kerner (1887) in a geographical sense, to designate the area of plants extending from the shores of the Black Sea to the Alps and Carpathians. This designation was used by subsequent investigators in an entirely different context, i.e. in a vegetational (steppe) ecological (heliophile), migratory (route) and in a historico-genetical respect (origin). The reason for these contradictions lies in the natural development of botanical science implying new plant-geographical approaches. The various authors include in the old singly existing term ‘element’ a variety of new plant-geographical aspects. The only suitable method of contending with these difficulties is to form adequate terms for each specialized aspect. In 1921 (Engler's Bot. Jahrb. 5: 598) Reichert suggested the use of the term ‘component’ for the designation of recent distribution, the term ‘migrant’ for migration of plants. He also desired to retain the old term‘ element’ for origin characterization in the sense of Engler (1878). In 1940 {Pal. Journ. of Bot., R.S., 8:183) Reichert proposed to revive the old name ‘type’ of Watson (1835) used by him for ecological description of plants for the same purpose. New proposals are now put forward for the characterization of a geographical natural area.