FOR the purpose of this paper the title of Physical Geography may be taken to mean the study of those processes of earth, air, and water which affect the conditions under which we live. Many geographers would limit the study strictly to the effects only of those processes, the bare facts of man's environ? ment, and leave the causes to specialist sciences. For those who believe in such rigid boundaries to the subject a laboratory for physical geography will have no appeal, but for those who feel that it is unnatural if not actually dangerous to divorce cause from effect, mechanism from final result, the idea of investigating the cause and nature of such processes must be welcome. We may therefore define the object of the physical geography laboratory as the study of field processes, usually in miniature, under conditions of close observation and control, with a view to ascertaining their mechanism, their stages, and their effects. There is nothing very original in such a development, which means only that the geographer is following the example of most other field sciences, which have for long been accustomed to take certain of their field_ problems into the laboratory for elucidation and experiment. Nor is it anything new that problems of physical geography should be submitted to experiment since many of the inquiries of civil engineers are essentially special cases of physical geography. This Society in particular has encouraged such work for a long time past, two excellent examples being Dr. Owens' investigation of the move? ment of material by water begun in the year 1907 (G.jf. 31 (1908) 415), and Major Bagnold's recent work on the movement of material by wind (G.J. 89 (1937) 409). That these papers have usually been presented by engineers and physicists rather than by geographers should be a spur to the idea that geo? graphers should undertake similar investigations. Some four years ago the Department of Geography at Cambridge decided to break ground in the direction of experiments in physical geography. We need not enumerate all the obstacles met with owing to scarcity of space and of funds, and not least to our own ignorance of how best to tackle the matter, for such obstacles are common to all new ventures. We should however like to record the assistance given by the University authorities and the satisfaction expressed by other scientific departments that physical geography was about to try and solve its own problems for itself. At the British Association meeting in Cambridge in 1938 the plans for such a laboratory were announced, the room was completed and in operation in 1939, and m normal times there should now be a crop of results of consider? able interest. Owing to the war the results have been but few and this paper must content itself With describing potentialities rather than actual achievements. A laboratory such as this is really a research unit, and requires researchers of ability and reasonable experience. Such cannot be produced at once, and there must first be a few years of teaching and sorting out of the research spirit amongst the taught. The apparatus is therefore used, at