346 HOW CROPS FEED. 
acted evidently in two or three days. Chloride of barium 
and nitrate of cadmium were slower in their effect. Chlo- 
rides of zinc and strontium at first, appeared not to react; 
but after twelve days, lime was found in the solution. 
Chloride of magnesium was still tardier in replacing lime. 
Four grams of powdered chabazite were digested with ' 
4 grams of chloride of sodium and 400 cubic centimeters 
of water for 10 days. The composition of the original 
mineral (1,) and of the same after the action of chloride of 
sodium (11,) were as follows: 
Silica, 47.44 48.31 
Alumina, 20.69 21.04 
Lime, 10.37 6.65 
Potash, 0.65 0.64 
Soda, 0.42 5.40 
Water, 20.18 18.33 
Total, 99.75 100.37 
Nearly one-half the lime of the original mineral was 
thus substituted by soda. A loss of water also occurred. 
The solution separated from the mineral, contained nothing 
but soda, lime, and chlorine, and the latter in precisely its 
original quantity. 
By acting on chabazite with dilute chloride of ammo- 
nium (10 grams to 500 c.c. of water) for 10 days, the 
mineral was altered, and contained 3.33 per cent of am- 
monia, Digested 21 days, the mineral yielded 6.94 per 
cent of ammonia, and also lost water. 
These ammonia-chabazites lost no ammonia at 212°, it 
escaped only when the heat was raised so high that water 
began to be expelled; treated with warm solution of pot- 
ash it was immediately evolved. The ammonia-silicate 
was slightly soluble in water. 
As in the instances above cited, there occurred but a 
partial displacement of lime, Eichhorn made correspond- 
ing trials with solutions of carbonates of soda and am: 
