668 



nitrate of potash lowers it very considerably. This opens np an inter- 

 esting- Hue of in vest! .nation. 



Th« relative rate of circulation of water within the soil depends upon 

 the amount of empty space in the soil; the number of grains per gram, 

 showiuf; the extent of subdivision of this empty spa<?e; the arrangement 

 of the soil grains; and tlic inlluence of the organic matter. 



The amount of empty space may be calculated from the weight of a 

 known volume of soil (conveniently taken with an ivfin cylinder (»f 

 known capacity driven into tin* ground) and tlie specific gravity' of tlic 

 soil itself. The api)roximate number of grains of sand and clay per 

 gram of soil can be calculated Irom the mechanical analysis. 



It has been suggested that the relative rate with which water circu- 

 lates witliin the soil can be calculated from the per cent of empty sjiace, 

 and the ai)pr()ximate number of grains per gram if the grains are 

 assume*! to have a mean symmetrical arrangement in ea«h soil and the 

 iiitbience of tlie organic matter is neglected. Tiie difference between 

 tiicdbserved rat*' »>f How determined experimentally and that calcnhitetl 

 from the nieclianical analysis of the soil should give an idea <»f the 

 arrangement of the s(»il grains, if the influence of the organic matter is 

 assumed to he the same in both sfuls. For examjtle, a tight, ahnost 

 impervious potleiy (lay has been slu)wn to have no more clay (45 per 

 «ent) than the subsoil of a fertile grass laml in a limestone forma- 

 tion. The relafi\<' rate of circulation calculated from the mechanical 

 analysi's of these two mat«Mials would be about the sanu», but a quan 

 tity of water passing through a given depth of the liuu'stone subsoil in 

 one huiulred minutes would requiie several days to jiass through the 

 ini|»ervious pottery clay. The dillerence in the rate of nu>vement, in 

 this case, \v<»uld l>e due to the dilleience in arrangen)ent of the soil 

 grains. The inlluence of the amount and coiulition of the organic mat- 

 ter on the rate of flow should foiin a separate line of investigation. 



T(t detrrniine the actual rafe ot<'irculalion of wati-r in lln' s(»il or sub- 

 soil in its natural position in a lield. a hole shonld be dug aiul the soil 

 an<l subsoil on one side n-niovcd to the de]>th at which the observa 

 tions are to be ina«le. A column of the soil or subsoil, - or more inches 

 square and 1 <u' 5 inches deep, is then to be carved out, and a glass or 

 metal frame a little larger than this and 3 or 4 inches deep is slipped 

 over the column and melted jiaralliu is then run in to fill up the space 

 between the s<til and the franu'. The soil is then struck off even with 

 the to]> ami bottom of the frame, a ]>iece of linen tied over the under 

 side, or, what is better, the frame can rest on sonu^ coarse sand <u' 

 gravel contained in a funnel, to ])revent the soil from falling out and 

 to juovide good drainage. A section <»f the franu' can then be i)laced 

 on the top and secured by a wide rublter baiul or otherwise, and the 

 time noted vhich is required for a quantity of wat<'r to ]>ass through the 

 satuiated soil. The initial depth of water over the soil must be the 

 sameiu all the experiments. The per cent of empty space mustof course 



