31 
tilizer, since this depends primarily upon the mechanical composition 
of the soil, the nature of the organic matter, and probably upon cer- 
tain factors which are at present unknown. However, within certain 
limits, the effect of adding a larger application of a salt only magnifies 
that of a smaller application. Tliis suggests that the measurement 
of the physical effect may be just as accurately, and possibly more 
accurately, determined than the chemical effect. The measurement 
of a normal application of fertilizer through a chemical analysis of 
the soil is practically impossible. 
Capillarity is diminished in clay soils by the addition of salts but 
increased in sandy soils. Also this property is more active in silts 
than in sandy or clay soils, being slowest in the latter. 
The percolation of water is most rapid in sandy soils and slowest 
in types the particles of which are most likely to swell. Fertilizers 
considerably increase the resistance to percolation. The theory that 
soils of greater capillary activity offer the least resistance to perco- 
lation of water does not apply to Hawaiian soils. 
Salts increase or diminish the size of the soil aggregates. This 
is of no small importance in the use of fertilizers. 
The cohesion of the soil particles in most instances is increased by 
the addition of salts. This is also true of the apparent specific 
gravity. However, there are too many exceptions to make any 
definite statement. 
The hygroscopic moisture is increased by the addition of salts, 
with but very few exceptions. 
The vapor pressure is lowered in most instances, but can not be 
explained from a consideration of the surface tension of the added 
salts. 
Acknowledgments are due and thanks hereby extended to Dr. 
W. P. Kelley for valuable suggestions and for interest shown 
throughout this investigation. 
ADDITIONAL COPIES 
OF THIS PUBLICATION MAY BE PROCURED FROM 
THE SUPERINTENDEXT OF DOCUMENTS 
GOVERNMENT PRINTING OFFICE 
WASHINGTON, D. C. 
AT 
5 CENTS PER COPY 
V 
