14 



EXPERIMENT STATION RECORD. 



as best for wheat; and in dry countries, such as Italy, those which 

 contain most ahimina. The sandy soil does not draw up or retain mois- 

 ture, which is not needed in the north, while the alumina conserves the 

 somewhat deficient moisture of the southern region. 



There is an evident relation between the quantity of water which 

 circulates in a plant and the dry matter elaborated. This relation is 

 as yet little understood. Deherain, repeating- the experiments of 

 Haberlandt, has found that 1 kg. of dry substance is elaborated by a 

 plant on an unmanured soil for every 680 kg. of water transpired by 

 the i)lant, and for every 220 kg. of water upon the same soil when 

 manured. Wollny has also studied this question on manured and 

 unmanured soils. Hellriegel has found that there exists an optimum 

 for the production of dry matter in ])lants, and that the j)roduct varies 

 according to tl»e amount of moisture. Representing by 100 the quan- 

 tity of water necessary for a complete saturation of the soil, Hellriegel 

 found that the production of dry matter in barley varied with different 

 water contents of the soil as indicated in the following table: 



l)ry matter in harlei/ on soil witli different water contents. 



The table shows that the optimum humidity is not the same for grain 

 and straw. These results show that in the case of two soils, the one 

 containing 30 per cent and the other 60 per cent of water, more grain 

 would be grown on the first plat than on the second. 



If experiment stations would conduct upon a large number of plants 

 the experiments described above, results of great practical value would 

 be secured. For example, with two dry soils containing from 10 to 15 

 per cent of water and capable of irrigation, what should be the method 

 followed'? Will it be necessary to continue a humidity of 60 per cent 

 during the entire period of growth? The author's experiments show 

 that there are many possible ways in which to secure a maximum 

 yield. To give water to a plant at the proper time after a slight 

 drought gives better results than maintaining a permanent optimum 

 humidity. Periods of relative humidity and drought are in general 

 very advantageous to plants, and the number of land jilants which 

 require a i^ermanent humidity for maximum i^roduction is very small. 

 It may be asserted as established that (1) for a given jdaut an inter- 

 ruption, however short, during a dry period proves very beneficial, and 

 (2) some i)lants which require a certain amount of humidity during one 



