THE RARER ELEMENTS IN SOILS AND PLANTS. LG 
aluminum content probably is due. Pine needles are exceptionally 
high in this element. 
5. There is no evidence that vanadium replaces phosphoric acid in 
its functions in the plant. Rubidium and caesium are apparently 
absorbed by the plant when they are present in the soil solution. 
The larger the amount present in the soil, the more will be absorbed. 
Lithium is apparently not acted on in this manner, for Nos. 12 and 
34 did not contain an exceptional amount of lithium, though the soil 
contained considerable lepidolite. 
6. With the exception of sulphur, chlorine, and manganese in some 
cases, there is no indication that the elements determined in the 
foregoing work, except those commonly used, need be considered in 
fertilizer practices. 
APPENDIX. 
ANALYTICAL METHODS. 
Since this work is based largely on analytical data, it seems proper 
that the methods employed should be stated with some detail. No 
satisfactory method for the complete analysis of plant ash is to be 
_ found in textbooks or bulletins. The problem is one of peculiar 
difficulty. The inorganic composition of plants nearly resembles that 
of a rock phosphate, except that there is a high content of magnesium, 
more iron and aluminum, and occasionally large amounts of man- 
ganese. Obviously some method which would remove the phos- 
phoric acid early in the analysis and leave no objectionable precipi- 
tant in solution would be the ideal procedure. The basic acetate 
method for the precipitation of the iron group had to be discarded 
on account of the tendency of the small amount of iron and aluminum 
phosphate to run through the filter. 
Treadwell and Hall have outlined a method whereby the phos- 
phoric acid is precipitated with molybdic acid in the usual manner 
and the iron and aluminum separated by ammonia and ammonium 
sulphide, which does not precipitate the molybdenum. Test analyses 
made on synthetic solutions showed that iron and aluminum were not 
completely precipitated by this treatment, and the method was aban- 
doned as unsatisfactory. 
The phosphoric acid may be removed by silver salts in slightly 
acid solutions, but the coprecipitation of iron renders this separation 
useless. 
The stannic acid made by dissolving tin in a nitric-acid solution 
containing a phosphate will precipitate the phosphoric acid as tin 
phosphate, and no tin remains in solution. Unfortunately, iron is 
precipitated with the tin, so the method could not be used. 
