128 BULLETIN 1059, U. S. DEPARTMENT OF AGRICULTURE. 
ignition, suffers a loss of water. The experience of the writers in- 
dicates that the sand filter devised by Grandeau, with also the paper 
filter used by Hilgard, will come nearest holding the colloids in the 
soil. Such as are likely to escape will have passed the filter by the 
time the acid treatment is complete. In the case of the coarser forest 
soils the addition of sand is wholly unnecessar}^. 
Alway suggests the recording of the humus ash percentage as 
well as the humus as a means of detecting the errors which commonly 
enter into this determination. 
CAPILLARY CONDUCTIVITY 
As has been frequently pointed out in the discussion of the mois- 
ture problems, the rate at which a plant is able to obtain water from 
the soil particles with which the roots are in actual contact may have 
an important bearing on the wilting coefficient for the whole soil mass, 
and may, in turn, depend largely on the facility with which the mois- 
ture travels from one soil particle to another when there is unequal 
distribution. Thus, a clean sand is generally understood to have the 
highest conductivity (whether because of the close contact between 
the particles or because of the clearness of the spaces between the 
larger particles, is not known) , while clay in the soil seems to impede 
this movement, probably because of absorption, and humus apears to 
retard the movement, possibly by breaking the contacts between the 
mineral particles. 
The whole subject of capillary conductivity appears to have been 
thrown into confusion in recent years by practical findings, especially 
in connection with the study of moisture supplies in the arid farming 
regions of the West. In brief, it has been found that in certain lo- 
calities the soil is never moistened to a greater depth than, say, 10 
feet, by precipitation; that the moisture which goes beyond the 
depth of ordinary crop roots is never brought toward the surface by 
capillary action, and hence is lost for practical purposes ; that fallow- 
ing with the object of storing moisture in the deep soil is therefore 
useless. 
Buckingham (116) and McLaughlin (132) have apparently made 
the most exhaustive studies of the movement of soil moisture ; and 
it may be said that these investigations confirm the practical con- 
clusion that when the mean moisture content is very low and the 
difference in moisture between two points is slight, the rate of move- 
ment from the moister to the drier point is negligible. These, of 
course, are the conditions to be dealt with as the wilting coefficient 
is approached, and it is somewhat relevant to remark that there 
is no evidence against the ordinary conception of capillary move- 
ment when the amount of moisture in the soil is considerable. It is 
true, however, that this movement is very slow upward — that is, 
against the force of gravity. 
