20 



EXPERIMENT STATION RECORD. 



It is not a question of storing up fertilizing matter iu tlie soil, but 

 of rapidly disseiniiiatiug it through the soil by means of the water. 

 Experiments by Voelcker ou the absorbent power of different soils on 

 dilute solutions of ammonia, phosphates, and i)otash confirm this con- 

 clusion. 



Ronna showed that meadows in England had been irrigated witli 

 water containing a large amount of lime with admirable results. It is 

 only necessary to avoid such waters as come from peaty or marshy 

 soils, since ordinarily they are charged with sulphate of iron. 



It is not necessary for us, in a general dissertation, to go into details 

 as to the quantity of water required for irrigating under each method 

 of culture. The quantities vary according to the meteorological con- 

 ditions of the regions under consideration. In the south of France 

 about 16,000 cubic meters of water per hectare are required for C 

 months' irrigation. Nevertheless, we are guided somewhat by expe- 

 rience, and give to tlie agriculturist as a basis the following general 

 curve based upon the results of experiments: 



Water. 

 Per cent 



00 



50 

 40 

 30 

 20 

 I 



-Saturation 



Fruiting Maturity 



Germination 



Ut k 



Ist.flowering 2ndf lowering 



Fig. 1. — Diagram showing water requirements of ]ilants at difterent stages of growth. 



This curve is intended to indicate for plants of ordinary habits of 

 growth the optimum amounts of water in the soil at diflerent stages, 

 the amounts being stated in percentages of saturation. It will be noted 

 that the curve is characterized by a regular alternation of humidity 

 and relative drought, as already exi^lained.' From the point of view 

 of legislation and cooperative societies for irrigation there is, at least 

 iu France, a serious difficulty in the use of this curve. The farmers 



iThe experimental basis upon which the curve rests has been discnased at leni^th 

 in a previous article by the writer, an abstract of which will be found in E. S. R , 

 7, p. 366, 



