Mackinder : The Advancement of Geographical Science. 71 



we have a certain number of provincial Geographical Societies, 

 but, with the exception of the Royal Scottish Geographical 

 Society in Edinburgh, they are situated in the great commercial 

 centres, and devote themselves rather to the spreading of a 

 knowledge of the lands beyond the seas than to the study of 

 local British geography. Here again I must take a partial 

 exception in the case of the Royal Scottish Geographical Society 

 at Edinburgh ; but I think that I have not misrepresented the 

 very valuable aims and work of the Societies at Southampton, 

 Manchester, Liverpool, and Newcastle, or of the branches of 

 the Scottish Society at Glasgow, Aberdeen, and Dundee. In 

 course of time the geographical teachers in our old and new 

 universities may no doubt come to our aid, but there are wide 

 areas of our country which have no university, or none suffi- 

 ciently developed as at present to afford a Chair in Geography. ' 

 For some time to come I see no agencies which can cover the 

 United Kingdom consistently with centres of geographical study 

 unless they are to be found in the Societies which you represent. 



Let me now give a first indication of the nature of the work 

 which I am proposing. Many of your societies have members 

 interested in botany, and in your publications there are not a 

 few valuable memoirs dealing with the distribution of plant 

 species. That, of course, was a very necessary study, but we 

 are now developing a different study, whose object is to ascer- 

 tain the distribution of what are known as plant associations. 

 For instance, in the twenty-first and twenty-second volumes 

 of the ' Geographical Journal ' you will find maps showing the 

 distribution of the plant associations of Yorkshire, which have 

 been compiled from the researches of Dr. William G. Smith and 

 others who have assisted him. Here you will see carefully 

 mapped by Bartholomew the distribution of the various moor- 

 land, woodland, and farmland associations. For instance, 

 under the head of moorlands you will find distinguished upon 

 the map the bilberry summits, the cotton-grass bogs, the heather 

 moors, the grass heaths, the natural pastures, and the lowland 

 wamps. In each of these associations there are several char- 

 acteristic plants, which occur together and very rarely apart 

 — a fact which is obvious to anyone who contrasts the trees 

 and undergrowth which constitute an oak wood with those 

 which constitute a beech wood. Primarily, of course, the 

 distribution of these associations is due to differences of climate 

 and soil, but also it must be remembered that the dominant 



1908 March i. 



