88 MICHIGAN ACADEMY OF SCIENCE. 



the values of the grains in earli group lying witbiu definite known limits 

 or depending n]»on certain definite conditions. 



By one method this se]taratit)n is accomplished by means of sieves 

 of definite mesh. These sieves differ in size ranging from three mm. 

 to one-tenth mm. 



By another method the soil is brought into sus]»ension in water by 

 shaking and stirring, and is afterwards allowed to settle for definite 

 l)eriods ranging all the way from ten seconds to forty-eight hours, and 

 then pouring off" the water with all soil still in suspension. In this 

 case the time allowed for settling measures the value of the soil grain 

 in any group. In another form of this method the settling is so timed 

 that with the use of micrometer and microscope the groups obtained 

 in the separation have their grains lying between definite measured 

 limits. 



By the third method the separation is accomjjlished by placing the 

 soil in currents of water of different but definite rates of fiow. In 

 this case the rate of flow is the measure of the value of the grains in 

 each group. But with this method also, with some forms of ajtparatus, 

 it is possible to so regulate the curients that the sizes of grains of each 

 group lie within certain definite limits. 



The most satisfactory methods in actual practice combine the use of 

 sieves with one or other of the water methods. By the method most 

 common in this country at this time the grades are made as follows: 



1. Gravel, with the size of grain ranging between 2 mm. and 1 mm. 



2. Coarse sand with the size of grain ranging between .1 mm. and 

 .5 mm. 



3. Medium sand, with the size of grain ranging between .5 mm. and 

 .25 mm. 



4. Fine sand, with the size of grain ranging between .25 mm. and 

 .1 mm. 



5. Ver^- fine sand, with the grains ranging between .1 mm. and .05 mm. 

 G. Silt, with the size of grain ranging between .05 mm. and .01 mm. 



7. Fine silt, with the size of grain ranging between .01 mm. and 

 .005 mm. 



8. Clay, with the size of grain ranging fi'om .005 mm. down to smallest 

 particle. 



Th(^ first three of these grades are nmde with sieves, the fourth 

 ] tartly by sieve and partly by gravitation or elutriation; and the last 

 four by gravitation or elutriation. 



But there are serious objections to the most satisfactory of these 

 methods. The time required is great at best, days if the clay is sepa- 

 rated from the silt. The soil must be ])repared by boiling or other 

 ecpially tedictus and often unsatisfactory way. Flocculation of the soil 

 particles must often be guarded against. 



The results are not satisfactory in all respects. After the soils are 

 separated into grou])s, there is no way of determining the average size 

 of soil grains in the gi-ouj»s nor in the total soil. There is, therefore, 

 no way of deteiniining the size of the individual grains nor of estimating 

 the total surface of the soil which is desirable in many of the problems 

 arising in soils. The actual ]»iactice is to consider the mean of the 

 extreme limits of any grou]> so se])arated to be the average size of the 



