April 17, 1908] 



SCIENCE 



629 



86.7 and 165.8, tbe last average containing but 

 four analyses. 



This table here referred to contains only 

 analyses of a single soil type known as Cecil 

 clay as it occurs in the vicinity of Statesville, 

 N. C, and instead of indicating that all soil 

 solutions have the same concentration, if it 

 proves anything, it shows that even one and 

 the same soil type, as mapped by the Bu- 

 reau of Soils, may carry solutions of phos- 

 phoric acid ranging all the way from 16 to 165 

 parts per million, if indeed not from 12 to 203 

 parts. 



But from what has been demonstrated re- 

 garding potassium nitrate from clean sand, 

 and indeed from the data presented in Bul- 

 letin 32, referred to above, there is every 

 reason to think that the first three-minute 

 washing to which these soil samples were sub- 

 jected should not be expected to give a solution 

 which would indicate the true concentration of 

 the fixed water films investing the soil gran- 

 ules and soil grains, and which must be 

 regarded functional in root feeding, and hence 

 the true difference between the extremes in 

 this series of soil samples should be expected 

 to be much greater than that observed. 



During the year of preliminary studies re- 

 garding the nature and concentration of soil 

 solutions preceding the collection of the data 

 presented in Bulletin 22 a very carefully 

 selected series of samples of soils was taken by 

 the vmter from 67 localities covering 41 of 

 the bureau's soil types, distributed in seven 

 states, each sample being always a composite 

 of four cores taken with the soil tube instead 

 of with the auger. The total water soluble 

 salt content in the surface foot of 66 samples 

 from these soil types were determined by the 

 gravimetric method, the residues being dried 

 at 110° C. before weighing. These results 

 showed an extreme variation ranging from 105 

 parts per million of soil solution, computed to 

 a moisture content of 20 per cent, of the dry 

 soil, to as high as 1,962 parts; and when this 

 series of determinations is arranged in an 

 ascending order and the analyses averaged in 

 groups of ten the concentrations stand 289.5, 

 418.1, 546.8, 673.7, 810, 950 and 1,530 parts 

 per million, and when these differences are 



shown with only three-minute washings in 

 water equal to five times the weight of the 

 soil, in its fresh, normal field condition, a liter 

 of the solution being evaporated in each case, 

 it is difficult to understand how such data, 

 originating in the bureau, should have been 

 ignored in drawing conclusions so at variance 

 with the experimental records. In the case of 

 the phosphoric acid for this series, which was 

 determined for each of the surface four feet 

 separately and the data published in Bulletin 

 26 of the bureau, the extreme range of phos- 

 phates is from 1.64 to 39.56 parts per million, 

 and when grouped and averaged in tens as in 

 previous cases, the results are 3.64, 6.65, 9.65, 

 12.43, 17.57, 26.73 and 35.56 parts per about 

 4,000,000 of dry soil. In Bulletin 26 there are 

 also given data from various sources regard- 

 ing drainage waters from England, analyzed 

 by Way, Prankland and Voelcker, all of which 

 show strong differences in the concentration 

 of soil moisture as it leaves natural field soil 

 by way of underdrains. It must be conceded, 

 therefore, that there is no good foundation in 

 observation or, indeed, from a priori considera- 

 tions, for the contention that all soil solutions 

 have essentially the same composition and con- 

 centration when viewed from the standpoint of 

 their function in plant growth. 



Soluble Salt Content of Soils in Relation to 



Yield of Crops. 



It is further contended, from data presented 

 in Bulletin 22, that there is no apparent rela- 

 tion between the yield of crops and the soluble 

 salt content of soils and the statement is quite 

 true when thus based. But it should be gen- 

 erally understood that were the converse of 

 this contention very strongly marked such data 

 as are collected and presented in Bulletin 22 

 must be expected to fail utterly to reveal the 

 fact. These are the facts regarding the data 

 used: (1) Assumed, not measured, yields of 

 crops form the basis of comparison ; (2) single 

 and isolated soil samples for determination are 

 taken to reveal field conditions, which they 

 seldom do closely; (3) much of the chemical 

 work was done outside the laboratory under 

 conditions prohibitive of the necessary accu- 

 racy; (4) duplicate solutions prepared from 



