560 PROCEEDINGS OF SECTION G2. 



wheat for 50 years on five of the plots : — 



Bushels. Difference. 



No manure .. .. .. .. .. ,. 13*1 



Complete Mineral Manure each year .. .. 14 "9 +1'8 



Ditto, + Ammonium Salts, equal to 431b. Nitrogen 24 '0 +9 "1 



„ 861b. „ 32-9 +8-9 



„ 1291b. „ 37-1 +4-2 



The results varied somewhat from year to year according to 

 the season, but the same general order prevailed throughout, so 

 that, I think, it is not too much to say that the main determining 

 factor in the production of wheat under Rothamsted conditions is 

 the supply of a sufficiency of readily available nitrogen. 



Anyone who assumed that similar results would be obtained 

 under widely different climatic conditions would find himself very 

 much at sea. This was forcibly brought home to the writer by the 

 results of the first accurate manurial experiments carried out in the 

 Transvaal, South Africa. At the Government Experimental 

 Farm at Potchefstroom, on a soil containing little more nitrogen 

 than is contained in the Rothamsted soil, there was no increase in 

 the yield of wheat, oats or maize due to the application of a nitro- 

 genous fertilizer, either used alone or in conjunction with other 

 fertilizers. Again, a soil containing only 0'075 per cent, of nitrogen, 

 near Pretoria, failed to respond to a nitrogenous fertilizer with 

 maize, and many similar examples from South African practice 

 could be given in the light of more recent experiments. 



In the wheat belt of New South Wales, which has a rainfall of 

 17 to 26 inches, very similar results have been obtained. Thus a 

 soil at the Experimental Farm at Bathurst, containing exactly the 

 same percentage of nitrogen as the Rothamsted soil, gives no 

 increase, and a soil from the Wagga Experimental Farm containing 

 0'065 per cent, of nitrogen, a very slight increase through the use of 

 a nitrogenous fertilizer. And a sandy loam soil occurring near 

 Cowra, with only 0^042 per cent, of nitrogen — a hopeless quantity 

 from an English point of view — has, during six out of the last 

 seven years, given an average crop of 16| bushels of wheat to the 

 acre, thus showing no great need for a nitrogenous manure. 



It is obvious, then, that a smaller proportion of total nitrogen 

 suffices to give a full crop in semi-arid Australia than in moist and 

 temperate England and that nitrogenous fertilizers do not have 

 such an effect as one would expect from English experience, and the 

 object of this short paper is to enquire into the reason for this. 



One explanation offered in the Transvaal was to the effect 

 that heavy rains washed soluble nitrogenous fertilizers into the 

 sub-soil out of the reach of plant roots ; but surely the holder of 

 such a theory must be ignorant of the fundamental principles of 

 soil physics and of the depth to which the roots of ordinary agri- 

 cultural plants penetrate in semi-arid regions. 



