INORGANIC COMPOSITION OF SOME AMERICAN SOILS. 7 
agreed fairly well and concordant results were obtained by both the 
preferred and alternative method outlined by Hillebrand. 
NICKEL AND COBALT. 
These elements were found in determinable amounts in the first 
soils analyzed. Later in the work only traces were found. The 
crucibles in which the fusions were made had been previously used 
in fusions of material which had been precipitated from an ammo- 
niacal liquid coming into contact with nickel plate. It is not impos- 
sible that some of the nickel reported in the first eight anaylses might 
have come from that previously absorbed by the crucible, although 
they were scrupulously cleaned by fusion with potassium bisulphate 
and sodium carbonate prior to the first soil fusion. A blank fusion 
after the first six analyses showed no nickel. Noyes, Bray, and 
Spear * have shown that nickel sulphide is not completely insoluble 
in cold hydrochloric acid (1:1 HC1), and since the method used calls 
for this separation, further work with this process was discontinued. 
COPPER. 
Copper was detected in the precipitate for nickel, cobalt, and copper. 
Since there is a possibility that there might have been contamination 
from the copper steam baths, by means of air currents carrying the 
oxidized copper, the presence of the element at this point in the ana- 
lytical operations is not to be taken as indicating that it is present 
in the soil. 
MOLYBDENUM. 
The test used for molybdenum was that recommended by Hille- 
brand. It is very delicate. 
LITHIUM, CESIUM, AND RUBIDIUM. 
Lithium was invariably present, giving stronger tests in some 
instances than in others. For the identification of this element and 
also for caesium and rubidium, a Hilger wave-length spectroscope was 
used. For lithium the line 6708.2 was used. The two doublets 
4593.3-4555.4 and 4215.6-4201.9 served to identify caesium and 
rubidium respectively. To determine the rare alkalies 10 grams of 
soil were fused with 10 grams of calcium chloride and 40 grams of cal- 
cium oxide for 4 hours. The chlorides of all the alkalies were 
obtained by the method of J. Lawrence Smith. To the solution of 
the chlorides about 0.05 gram of platinic chloride was added, the 
solution stirred, and evaporated to pastiness. The unchanged 
chlorides were rapidly dissolved in a minimum amount of hot water, 
the platinum salts washed onto a small carbon filter, dried and re- 
1 A System of Qualitative Analysis, p. 78. 
