REDISTRIBUTION OF MANKIND- — DICKSON. 567 



the protection of such resources pending investigation is already 

 much needed. It is worth noting that the question may often be 

 closely related to the development and transmission of electrical 

 energy from waterfalls, and the two problems might in such cases 

 be dealt with together. Much may be learned about the relation of 

 water supply to distribution of population from a study of history, 

 and a more active prosecution of combined historical and geograph- 

 ical research would, I believe, furnish useful material in this connec- 

 tion, besides throwing interesting light on many historical questions. 

 Continued exchange of the " north-and-south " type, and at least 

 a part of that described as " east-and-west," gives permanence to a 

 certain number of points where, so far as can be seen, there must 

 always be a change in the mode of transport. It is not likely that 

 we shall have heavy freight-carrying monsters in the air for a long 

 time to come, and until we have the aerial " tramp ", transport must 

 be effected on the surfaces of land and sea. However much we may 

 improve and cheapen land transport, it can not in the nature of 

 things become as cheap as transport by sea. For on land the essen- 

 tial idea is always that of a prepared road of some kind, and, as 

 Chisholm has pointed out, no road can carry more than a certain 

 amount; traffic beyond a certain quantity constantly requires the 

 construction of new roads. It follows, then, that no device is likely 

 to provide transport indifferently over land and sea, and the sea- 

 port has in consequence inherent elements of permanence. Im- 

 proved and cheapened land transport increases the economy arising 

 from the employment of large ships rather than small ones, for not 

 only does transport inland become relatively more important, but 

 distribution along a coast from one large seaport becomes as easy as 

 from a number of small coastal towns. Hence the conditions are in 

 favor of the growth of a comparatively small number of immense 

 seaport cities like London and Xew York, in which there must be 

 great concentration not merely of work directly connected with 

 shipping, but of commercial and financial interests of all sorts. The 

 seaport is, in fact, the type of great city which seems likely to in- 

 crease continually in size, and provision for its needs can not in 

 general be made from the region immediately surrounding it, as in 

 the case of towns of other kinds. In special cases there is also, no 

 doubt, permanent need of large inland centers of the type of the 

 "railway creation," but under severe geographic control these must 

 depend very much on the nature and efficiency of the systems of land 

 transport. It is not too much to say (for we possess some evidence 

 of it already) that the nimiber of distinct geographical causes which 

 give rise to the establishment and maintenance of individual great 

 cities is steadily diminishing, but that the large seaport is a per- 

 manent and increasing necessity. It follows that aggregations of 



