June 19, 1891.] 



SCIENCE. 



343 



pied tky the soil. The light saudy sands, therefore, are light 

 only in texture, for a cubic foot of sand weighs about one 

 hundred and ten jDounds, while an equal volume of clay 

 weighs seventy-five pounds. 



The water of the soil has to move in this empty space, and 

 the relative rate of movement will depend upon how many 

 particles there are in the soil, for this will determine the num- 

 ber and size of the spaces between the particles in which the 

 water will have to move. 



The soil particles vary in size from about 2 millimetres in 

 diameter to about .0001 of a millimetre, which is near the 

 limit of microscopic vision. The coarser particles are called 

 sand, while the very finest particles are known as "clay." 

 We cannot emphasize this point too strongly, that clay differs 

 from saud only in the size of the grains. The particles of 

 <5lay are hard and compact as sand, are composed largely of 

 quartz, and they have themselves none of the inherent sticki- 

 ness associated with clay in mass. 



The plasticity of moist clay and the hardness of dry clay 

 in mass, as distinguished from the looseness and incoherency 

 of sand, is due to the fact that the clay has a vastly greater 

 number of particles in a unit mass than sand has, and as each 

 grain touches the surface of six or eight adjacent grains, there 

 are many more points of contact for surface attraction to act 

 and bind the mass of clay together. 



The approximate number and size of the particles may be 

 found or calculated from the mechanical analysis of a soil. 

 The mechanical analysis consists in separating the particles 

 into eight or ten or more grades whose diameters range be- 

 tween rather narrow limits by sifting and subsidence in 

 water. 



The mechanical analysis in its simplest form, as devised 

 by Nobel and adopted some years ago by the Society of Agri- 

 cultural Chemists of Germany, consists in boiling the soil for 

 some time in water, to disintegrate any lumps, and jjlacing 

 it in the first of a series of conical-shaped vessels having a 

 capacity respectively of 1, 8, 27, 6i. A stream of water is 

 let in which carries the finer particles over into the next suc- 

 ceeding larger vessel, where, the motion of the water being 

 slower, grains of somewhat smaller size may settle, and so on. 

 Many small grains are, however, carried down with the large 

 ones, and Hilgard has improved on this by having a paddle 

 revolving at a high speeed in a porcelain cup, which keeps 

 the soil thoroughly agitated. From here the mixture rises 

 into a wide tube sufficiently high so that large grains thrown 

 up by the current of the paddle will not go over. When the 

 water comes over clear the receiving vessel is changed, and 

 the velocity of the water is increased so as to carry over 

 grains of a larger size. Johnson and Osborn have simplified 

 this in the following method. The soil is gently rubbed up 

 in a mortar with a rubber pestle with repeated quantities of 

 water, until the water, after standing a moment over the soil 

 in the mortar, is perfectly clear and all grains smaller than .05 

 of a millimetre have been removed, as shown by microscopic 

 measurements. The coarser grains are then sifted in a series 

 of sieves. 



The turbid liquid is allowed to stand until all particles, 

 larger than a certain size, have settled, as shown by micro- 

 scopic measurements on a drop of the liquid removed with a 

 rod or (bbe. The turbid liquid is poured off into another 

 heaker w settle, and the contents of the first beaker is stirred 

 up with a fresh quantity of water, and the settling continued 

 until all particles, smaller than a certain size, are removed, 

 and so on for the several grades. The separations are finally 

 dried and weighed. 



The following table gives the mechanical analysis of two 

 markedly different types of soil: — 



From the results of the mechanical analysis, the approxi- 

 mate number of particles in the soil can be calculated from 

 this formula: — 



7r(d)^2.65 -;- total weight of soil. 

 tJ 



Where a is the weight of each group of particles, d the mean 

 diameter of the particle in the groups, and 2.65 taken as the 

 specific gravity of the soil. 



From this and the weight of a unit volume of soil, the 

 number of particles on a unit area of surface can be calcu- 

 lated. 



(.! 



No. particles in 1 cc 



y 



There will evidently be one space or opening into the soil 

 for every surface grain. If the grains have a symmetrical 

 arrangement the mean size of these spaces can be calculated 

 from the formula: — 



Where r is the radius of the space, V the total volume of 

 all the space, N the number of spaces on a unit area, and 

 L the depth of soil. 



The circulation of water through the soil will depend upon 

 the size of these spaces, and not in any simple ratio either, 

 but according to the fourth power of the radius multiplied 

 by the number of spaces. You will bear in mind we are not 

 trying to establish absolute but i-elative values. 



Here are ten tubes, each with a radius of three units, and 

 here is one single tube with a radius of ten units, having the 

 same capacity and area of cross section of the ten tubes. If 

 they were exceedingly small capillary tubes water would 

 flow through the single large tube about twelve times faster 

 than through the ten tubes. So it is in the .soil. If we as- 

 sume that there is the same amount of empty space in a clay 

 soil as in a sandy soil, there are at least ten times the num- 

 ber of spaces in the clay soil for the water to move through, 

 and the movement is very much slower than in a sandy soil. 

 Clay has no inherent property of absorbing and holding 

 moisture not possessed by sand, as popularly supposed, the 

 difference being due entirely to the number of particles per 

 unit mass. 



