628 



SCIENCE 



[N. S. Vol. XXVII. No. 694 



eideration of this question to recognize the 

 fact that neither the water solutions decanted 

 from soil samples, the water filtered through 

 soil, nor the solutions separated from field 

 eoil by either natural drainage or a centrifuge 

 represent either the relative composition or the 

 concentration of the portion remaining behind 

 in contact with the soil. The only conclusion 

 which is warranted, based upon existing pub- 

 lished data regarding the concentration of soil 

 solutions as they exist under field, plant house 

 and pot culture conditions, is that they are 

 certainly an undetermined amount stronger 

 than would be indicated by the observed con- 

 centrations of solutions recovered thus far 

 from, soils under such cultural conditions. 



It had been demonstrated in the laboratory 

 of the Bureau of Soils, before the data pre- 

 sented in Bulletin 22 had been obtained, that 

 when a coarse, thoroughly acid-cleaned and 

 washed sand known to contain 84.71 parts per 

 million of NO3 in so soluble a form as potas- 

 sium nitrate was washed three minutes in 

 twice its weight of water, only 73.7 per cent, of 

 the actual amount present was recovered in 

 the first washing, and after ten consecutive 

 eimilar washing's there was still retained in 

 the moisture films closely investing the sand 

 grains 18.02 per cent, of the original amount 

 given to the sand, which amount was subse- 

 quently recovered after rendering the sand 

 water-free. After the sixth washing, in this 

 experiment, the strength of successive solu- 

 tions became constant at .11 part per million 

 of NO3, the concentration standing as below : 



Sehreiner and Failyer, in Bulletin 32, Bu- 

 reau of Soils, have since published similar 

 results, working with very soluble phosphates 

 and potassium chloride in connection with 



several soils. The simplest explanation for 

 results like these is to suppose that these salts 

 are retained in solution within the stationary 

 films investing the sand and soil grains and 

 that the salts recovered by successive wash- 

 ings after the concentrations become constant, 

 measure the rate of outward diffusion from^ 

 the stationary film into the free water. But 

 the Bureau of Soils holds (and presumably 

 the authors of Bulletin 32) that the fact that, 

 after prolonged leaching of soils with pure 

 water, the filtrate appears to attain a constant 

 concentration, furnishes direct proof in sup- 

 port of the contention that all soil solutions 

 have a concentration sensibly the same, the 

 thought being that certain amounts of the 

 salts experimented with had been absorbed by 

 the soils, which were later redissolved by the 

 excess of water on its passage over the soil 

 grain surfaces. It is to be observed, however, 

 that in the case of the solutions and soils re- 

 ported upon by Sehreiner, the salts retained, 

 even in the case of phosphates, had not be- 

 come sufficient to make the retained water a 

 saturated solution under conditions not 

 affected by the soil surface action. It must, 

 therefore, be admitted, we think, that the data 

 of Bulletin 32 lends little support, such as is 

 being eagerly sought, in defense of the views 

 of Bulletin 22, to wit, that in all cultivable 

 soils the moisture is a nutritive solution hav- 

 ing sensibly the same concentration. 



So far is this view from being a demon- 

 strated fact that if the data presented in 

 Bulletin 22 in support of it be admitted in 

 evidence quite the reverse proposition must be 

 counted proved by nearly every table there 

 presented. To illustrate: The data on pages 

 31, 32 and 33 show a range of soluble phos- 

 phoric acid varying from 12 parts per million 

 of soil moisture to 203 parts per million, sup- 

 posing the water content in each soil sample 

 to be 20 per cent., computed on the dry soil. 

 That this wide range in concentration is not 

 due to exceptional cases is made clear by 

 arranging the data of this table in an ascend- 

 ing series ; averaging in groups of ten to re- 

 duce errors and unusual irregularities, we get 

 the following series of concentrations : 16.8, 

 28.8, 37.2, 41.7, 45.4, 50.3, 54.6, 59.8, 64.3, 70.6, 



