| aa 
THE RARER ELEMENTS IN SOILS AND PLANTS. 23 
The temperature of the furnace is controlled through the action of 
the nickel rod which is placed inside a quartz tube. The quartz tube 
was fastened at the rear of the furnace to the inside of a brass piece 
by means of glycerine and litharge cement, after the end of the tube 
had been wound with asbestos cord. The other end of the lever con- 
nected with the nickel rod makes and breaks the current operating 
the relay. A condenser is connected with opposite sides of the gap 
to reduce arcing. The platinum-tipped contact pieces are also im- 
mersed in oil, which further reduces arcing tendencies. An ammeter 
and rheostat are connected in series with the furnace. When started 
with a current of 12 amperes (at 220 volts) the furnace requires about 
an hour and a half to reach a temperature of 500° C., after which the 
temperature may be kept constant with a current of 7 amperes, 
which is easily handled by the relay. 
The time required for ashing of different plants varies consider- 
ably. Three hours is generally sufficient. The contents of the ashing 
dish were turned to expose unburned portions when necessary. 
Many times one incineration was suflicient for complete ashing, but 
generally all plants high in silica required a second heat treatment 
after leaching with water. The silica apparently forms a slag with 
the alkalies and protects enmeshed particles of carbon from the action 
of the oxygen. 
It was found more convenient to determine the sodium and potas- 
sium in the same portion that was taken for the major constituents. 
The ash was dissolved in hydrochloric acid, with repetition of the 
ignition and extraction of the residue if any, and the filtrate made up 
to definite volume. An aliquot corresponding to 1 or 2 grams was 
taken for the alkalies. This was evaporated to dryness to free from 
the excess of acid. Clear barium hydroxide solution was added in 
slight excess and the solution filtered. All interfering elements are 
eliminated by this process, and from this point on they Oey pro- 
cedure is followed. 
For lithium, caesium, and rubidium 20 grams of the dry plant were 
taken. This amount was carefully ashed, the ash dissolved in hydro- 
chloric acid, and the excess of acid evaporated off. The magnesium, 
phosphoric acid, etc., were precipitated by lime, the lithium and 
sodium separated from the other alkalies by platinum-chloride, and 
the caesium and rubidium fractionated from the potassium in the 
same manner as outlined in Bulletin No. 122 of the United States 
Department of Agriculture. The determinations were made spec- 
troscopically according to the method of Gooch and Phinney! with 
a Hilger wave-length spectroscope. Lithium was identified by the 
line 6708.2 and caesium and rubidium by the doublets 4593.3—4555.4 
and 4215.6—4201.9, respectively. 
1Am. J. Sci., 44, 392 (1892). 
