1917] Lipman-Geriche : Smelter Wastes and Barley Growth 499 



irrigation water and the mode of its application were most 

 desirable under the circumstances and were such as to preclude 

 losses of water and salts by percolation and drainage. 



Barley {Hordeum vulgare) was the crop grown throughout 

 all series of experiments. The variety employed was a selected 

 and vigorous strain of Beldi. Three crops were grown in suc- 

 cession on the humus clay adobe soil, the first and third crops 

 being produced in the period between September and January 

 of the years 1912-1913 and 1913-1914 respectively, and the sec- 

 ond crop between March and June, 1913. Only two crops were 

 grown on the non-humus clay adobe soil, in periods correspond 

 ing to the last two for the humus clay adobe soil. One crop only 

 was grown on the blow-sand soil. 



At the time of harvest, which was carried out when the grain 

 was thoroughlj^ mature, the plants were cut as close as possible 

 to the ground. The total crop thus obtained was placed in paper 

 bags and dried until the weight was constant. Then the weights, 

 separately, of grain and straw were determined. At the same 

 time the soil in tlie pots was thoroughly worked over to obtain 

 the roots produced in every case. In some instances nitrification 

 studies on the soil were made, and also determinations of the 

 amounts of salts remaining behind in the soils after harvest, 

 and the amounts taken up by the crop. Enough of these anal- 

 yses, as well as of nitrogen determinations in the grain, were 

 accomplished to ascertain the tendency of these conditions in 

 the plants and soils studied. 



RESULTS OF THE EXPERIMENTS 



It will undoubtedly be of interest to our readers to learn first, 

 from the results of our experiments, something of the appear- 

 ance, height, tillering, color, and similar observations on the 

 growth of the barley, and later the yields obtained, composition 

 of the grain, and changes in the soil. The following general 

 statements may therefore be made at this point with respect 

 to the first class of data obtained through the experiments. The 

 different features will be considered separatel3^ 



