——- 
TRANSACTIONS OF SECTION C. 365 
areas where, as in the case of the Kara Boghaz of the Caspian, arms of the 
sea are partly cut off and subjected to desiccation. 
Although it is dangerous to generalise from this single instance of Rajputana, 
in spite of its striking and conclusive character, the observations made in that 
region are quoted as an instance of an attempt to check by definite quantitative 
tests general mental impressions of geological dynamics in desert regions. The 
object of this communication is mainly to urge the further institution, where 
practicable, of such tests of current theories regarding the physiographic 
phenomena of arid lands. 
Professor W. M. Davis: My interest in the subject proposed for our discus- 
sion comes from an endeavour to systematise the study of land forms, so that 
a well-trained explorer shall be aided in making accurate and complete observa- 
tions of the ground, and in preparing afterwards for readers as expert as 
himself a complete and intelligible record of his observations. It would be 
comparatively easy to reduce such a description to simpler or shorter form for 
more elementary or more popular use; but it would be impossible to expand 
a short elementary account intended for beginners, or a popular account 
intended for general readers, into a detailed monograph intended for experts. 
The advancement of geographical science will therefore be best promoted by 
striving to develop a mature thoroughgoing method for the observation and 
description of all kinds of land forms, including those of deserts. 
Much assistance has been given to the study of land forms in general by 
working out their evolution as dependent (1) on their structure, (2) on the 
erosional process that works upon them, and (3) on the stages which the forms 
produced by the work of process on structure pass through, from the initial 
stage introduced by the movement of a land mass into a new attitude, to the 
ultimate stage when the process concerned has done all its work. 
If we classify what has already been accomplished in this direction with 
respect to the erosional processes involved, it appears that the theoretical 
sequence of changes determined by the action of ordinary or normal processes 
on various structures has been worked out with encouraging success, and 
verified by confrontation with many examples of actual forms. The explana- 
tory method of describing land forms, based on this theoretical sequence, is 
now employed by a number of geographers. The same is true of marine erosional 
processes and of solutional processes. It is less true of glacial processes, though 
much good progress has been made in that division of the general subject. 
With regard to arid processes, theory has outstripped observation; hence the 
observational study of deserts is much to be desired as a means of testing, 
correcting, and extending the theory of arid erosion. The difficulty with the 
“descriptions of desert forms hitherto published is that they are so largely 
empirical and so incomplete that it is impossible to translate them into the 
phrases of rational or explanatory physiography. Hence what we now need 
is, the exploration of deserts by trained students, well informed regarding 
modern physiographic theories. . 
Let me illustrate this by a special case. The theory of the evolution of desert 
forms includes a stage in which a lower basin is about to capture the centripetal 
wash of a neighbouring higher basin; and another stage in which such a capture 
has recently taken place. The significant characteristics of each of these two 
stages, as well as of many earlier and later stages, have been defined with 
sufficient detail to make their recognition easy, provided that the observer is 
familiar with them; but it would be as unlikely that an observer untrained in 
physiography would see and describe the essential features of these stages of 
desert forms as that an observer untrained in botany would see and describe 
the essential features of plant forms. If one looks through various accounts of 
desert exploration, it is usually impossible to determine whether actual 
examples of imminent or of recent basin captures—or of any other special 
features of desert evolution—actually occur, 
The most helpful suggestion that I can offer in this connection is that the 
effort should be made to refer every element of desert topography first to 
its proper place with respect to the surrounding contemporary elements in the 
general working of the processes of desert erosion, and, second, to its proper 
