Skimk.muer 11. I!i0:i.] 



SCIENCE. 



343 



Oiii' fact is pruveil abundantl.v l)y a study 

 of the flora of tlie serpentine barrens, and 

 that is that the chemical character of the soil 

 derived from a disintegration of the serpen- 

 tine i)lays an iniiniportant part in the dis- 

 tribution of the plants mentioned. The dis- 

 tribution of such species is due rather to the 

 physical conditions of the soil, especially with 

 reference to water conductivity and water 

 storage capacity (edaphic conditions). The 

 variation in the character of the plant asso- 

 ciations described above is in the main due 

 to the character of the soil. If the soil is 

 present as a well-marked surface layer, then 

 tree associations are found; if on the other 

 hand the rock is exposed, herbaceous associa- 

 tions are the rule. The surface layers of 

 serpentine rock are broken by weathering into 

 angular fragments, which, lying loosely to- 

 gether, permit the percolation of the rain 

 water down into the seams of the underlying 

 rock. Such exposures, therefore, support 

 plants that have adapted themselves to living 

 in dry situations and have structural arrange- 

 ments which prevent a rapid loss of water. 

 John" W. Harsiiberger. 



I'xiVKitsiTV OF Pennsylvania. 



TlIK AMOIXTS OK REAnil.V WATER SOI.IBLE SALTS 

 FOUND IX SOILS UNDER FIELD CONDITIONS. 



Ix the investigations of the Division of Soil 

 ilanagenicnt, in the Bureau of Soils, relating 

 to the influence of soil moisture in crop pro- 

 duction it has been found essential to take 

 into consideration not only the varying 

 amounts of available moisture in the soil but 

 also the readily water soluble salts which this 

 moisture carries in solution. 



The sensitive and rapid methods which have 

 been devised or adapted for this work enable us 

 to determine the K. Ca, Mg, XO„ 11 PO., SO,, 

 CI, IICO, and SiO, in the soil witli an ac- 

 curacy of duplication ranging usnall.y from 

 one to five parts per million of the dry 

 weight of the soil examined and with rapidity 

 such that eight men are able to complete the 

 nine sets of determinations on twenty sam- 

 ples daily between 9 a.ji. and 4 p.ji. 



As these methods are now used in our soil 

 investigations, those for the K, CI and HCO, 



liiive been devised and adapted under the di- 

 rection of Dr. F. K. Cameron ; that for NO, 

 by A. K. Whitson of Wisconsin and the writer; 

 tliat for HPO. and SiO, by Dr. Oswald 

 Schreiner; those for Ca and Mg by Dr. 

 Sehreiner and W. S. Ferris, and that for SO, 

 b.v J. O. Belz. The clear soil solutions for ex- 

 amination are obtained by using the effective 

 filter devised by Dr. Lyman Briggs. 



After extended observations it has been 

 found that to recover the maximum amount 

 of the readily water soluble salts which are 

 present in the soil it is necessary to first render 

 the sample water free by drying at a tempera- 

 ture of 110° to 120° C, as soils are dried for 

 moisture determinations. Mr. J. O. Belz and 

 the writer found, for example, that after ten 

 times washing 50 grams of a coarse, clean sand 

 containing 4.125 mg.of potassium nitrate, that 

 the same sample oven dried after having been 

 ten times washed in 100 c.c. of distilled water 

 yielded when worked in the disulphonic acid 

 a large additional amount of nitrates. Our 

 actual figures are given below, where from 50 

 grams of sand we recovered : 



By Ist Hashing of three minutes. . . . ;j. 12100 mg. 



" 2(1 " - ■• " '• 32840 " 



" .'id " '■ " " 04515 " 



" 4th " " " " 01736 " 



'• ;">th " •■ " '• OI.-JSO " 



" Gth " '• " " 01280 " 



" 7th " •' " " 01109 " 



" 8th. " " " " 01100 " 



" 9th '• '• " " 01100 '• 



" 10th " " " '• 01101 '• 



After drying 7(1200 " 



Total recovered 4.:U.'>.>1 " 



Amoiuit present 4.12.J00 " 



These observations were made in February, 

 1902. Later in the season, in September, we 

 made an examination of thirty-two samples of 

 soil, representing eight soil types, determining 

 the amounts of NO,, SO., IIPO., HCO„ CI 

 and SiOj which could be recovered by washing 

 100 gTams three minutes in 500 c.c. of dis- 

 tilled water as they came fresh from the field, 

 and again by washing in the same manner 100 

 grams of the water free sample direct from 

 the oven. 



