of its comparative uniformity in this respect, is really uniform neither 

 in climate nor in soil, and these differences are, in all probability, 

 sufficient to meet the diverse physiological needs of the various types. 

 It is more than probable that it will be found, in fact it is found, 

 that the type which is best suited to one set of conditions will not be 

 the one best suited to a second ; and it will thus be possible to de- 

 marcate type tracts in each of which the cultivation of that type 

 which responds best to the local environment can be encouraged. 

 I do not overlook the work that has already been done in this direction ; 

 the point is appreciated, notably in the case of Ashmuni in Upper 

 Egypt, but I am inclined to think that a large scope still remains for 

 work in this direction on a systematic and predetermined plan. 



Such investigation is definitely agricultural, and it may, even so, 

 be found that though the yield of B may be increased relatively, 

 and even exceed that of A in certain areas, yet that increase still leaves 

 the product ax greater than by, still leaves, that is, A the more 

 profitable crop. To meet such a case, the possibilities of the second 

 method of equalizing the money value must be investigated. 



It has been repeatedly pointed out that commercial cottons of 

 Egypt are impure : that is, that they consist of a mixture of a larger 

 or smaller number of types, together with a considerable admixture of 

 crosses between these types. It is to this admixture that the 

 deterioration of quality, so noticeable in most Egyptian cottons, and 

 even that striking phenomenon of the definite limit to the life of any 

 particular variety, is attributable. In such a mixture the total yield 

 is clearly an average one. certain of the component types yielding more. 

 and others less, than the average. Also certain of the types will be 

 better, and other less, adapted in their physiological reaction to the 

 local environment. The replacement of the mixture by those 

 component types, or that component type best adapted to the 

 environment, can have only one result, the raising of the unit yield 

 from that of the mixture to that of the best of the component types. 

 This method is what is commonly termed selection, and we may 

 conveniently consider here the third point raised in discussing the 

 economical aspect, the need for uniformity, as well as the question of 

 uniformity just referred to. 



I have had occasion to observe that lint diversity, such as exists 

 between the different classes of cotton, is, in practice, associated with 

 a different physiological response to environmental conditions. The 

 converse is equally true ; mixtures such as compose and yield the main 

 classes of cotton contain types which not only exhibit these physio- 

 logical differences but also lint differences. Selection within the limits 

 of the present accepted classes thus is not limited in scope to the 

 isolation of what we may term physiological types. It may and should 



