II DEVELOPMENT AND ENVIRONMENT 45 



aflFairs was strikingly shown to the author by Mr. J. I. 

 Craig, who computed that on this basis the whole Nile 

 flood would be lost through the transpiration of the cotton 

 crop alone ! 



It may be possible, however, to obtain data from plants 

 growing under nearly normal field conditions, by under- 

 pinning large masses of earth to form concrete tanks in 

 the middle of a field of cotton, and determining the fall 

 of the water-table in these tanks from day to day. The 

 method is open' to many objections, but it lends itself to 

 automatic recording, and if regarded in part as a problem 

 in soU physics, the results should form a near approxima- 

 tion to the truth. It is not however suitable for examina- 

 tion of the diurnal variation. 



For the problems of diurnal change it is usual to 

 employ the Potometer, which compares the velocities of 

 intake of water by means of a tube intercalated in the 

 cut stem. This method has not been persuaded to work 

 in Egypt owing to the severity of the sunshine effect, 

 which promptly kills the severed stem. 



The only method which holds out any prospect of 

 utility is the least obviously praetical of any. It consists 

 in cutting off portions of plant tissue of known area, and 

 weighing them repeatedly on a Joly spring balance as 

 soon as possible after cutting. The weighing is done in 

 the field, usually under difiiculties due to the wind, and 

 the balance is set up so that the tissue shall remain in 

 the same position which it occupied before being cut off. 

 Even whole plants may be treated in this way and 

 measured up with a planimeter afterwards. The curve 

 of weight-diminution in time is then plotted backwards 

 to the moment of cutting, and its value at that moment 

 is assumed to be the true transpiration rate. This is then 

 converted for purposes of comparison into milligrammes 

 per square centimetre per minute. 



The few determinations yet made in this way have 



