192 FE. W. Hilgard— Objects and Interpretation of Soil Analyses. 
I have attempted to make allowance for the cases of dilu- 
tion, as above noticed, by combining the results of the mechan- 
ical with those of chemical analysis. In the investigation made 
by Dr. Loughridge, of the several sediments obtained in the 
mechanical analysis of the typical soil above referred to—see 
this Journal, Jan. 1874—it appeared that plant-food practically 
ceased to be extracted from sediments exceeding 5™ hydraulic 
value; and in re-calculating the percentages of soils of the same 
general derivation, after throwing out the coarser sediments, we 
often find very striking approximations to identity of percentage 
byi- 
composition, as well as of proportionately inter se. It is o 
by the method heretofore described by me, and not in accord- 
ance with that of the German experiment stations), as an almost 
indispensable aid in judging fully of the agricultural peculiari- 
ties of soils, especially when these cannot be personally exam- 
ined in the field. 
e concentration of the available portion of the plant-food of 
soils in their finest portions is almost a maxim already, scarcely 
needing the corroboration afforded by the investigation of Dr. 
Loughridge above quoted. A “strong” soil is invariably one 
containing within reach of the plant a large amount of impal- 
pable matter, although the reverse is by no means generally 
true. Striking corroborations of this maxim are afforded by 
the steady increase of certain plant-food percentages in the 
deposits of streams as we descend, and the proverbial richness 
of “delta” soils is exactly in poin : 
“Thigh” and “low” percentages and their interpretation.—1 will 
now state, as concisely as possible, some of the main points + 
consider as substantially proven by the comparisons of soil 
analyses made upon the uniform plan outlined above. e 
when 
the significance of the occurrence of certain trees, esteem 
the ‘old farmer” as certain signs of a productive soil, 1s Once 
the term, however, in a somewhat different meaning from an 
usually given it. That is, I find that in order to manifest 1tsew 
over: 
