On the Food of Plants. 
541 
ant office, that of aljstr.ictino- ammonia from llio air: oxide of iron, 
an abundant constituent of clay, having a gr(!at attraction for that 
substance. In the same manner, gypsum or sulphate of lime 
performs the double function of serving directly as saline food, 
and fixing the carbonate of ammonia of the rain-water, which 
would otherwise in great part evaporate from the surface with the 
water itself. 
Manures, properly so called, may be divided into azotized and 
saline, or those which give nitrogen, and those which restore to 
the soil the mineral nourishment withdrawn from it by plants : 
there is a third and most important class, to which urine belongs, 
the members of which perform both services. 
Among the saline manures we have gypsum, already mentioned ; 
wood-ashes, valualjlefor potash, salts, and phosphates; soot, valu- 
able for its sui])hate of ammonia, and perhaps for its carbon, which, 
Irom an experiment of Davy, seems to undergo slow oxidation by 
exj)osure to air and water ;* &c. 
Tlie only purely azotized manure is the ammoniacal liquid of 
the gas-works ; a very variable mixture of sal-ammoniac and sul- 
phate of ammonia, with generally a large proportion of sulphuret 
of ammonium and small quantities of other substances. This 
liquid has been tried by many persons as a manure for corn, but 
with vei'y variable success, as may be expected from its indefinite 
nature. In a little experiment which came under my own ob- 
servation, wheat so treated grew up in the rankest and most lux- 
uriant manner, presenting a deep green colour strongly contrast- 
ing with another portion of the same corn to which none of the 
manure had been applied. The best mode of using this sub- 
stance will certainly be to reduce it to an impure ammoniacal 
salt, sulphate, or chloride, so as to get rid of the sulphuret, sul- 
phite, &c., which can hardly fail to be prejudicial to vegetable 
life, and then to apply such salt scattered over the ground in a 
sparing manner : always on the supposition that the compounds 
mentioned have been shown to be capable of assimilation by the 
• Analysis of coal and coke ashes : — 
StafToidsliiie N'cwcisllc ' Staflbvdsliirp 
Coal. Coal. Coke. 
Sandy matter and unbunit charcoal . . 04-0 87 -C 7G-8 
Oxides of iron and manganese, alumina, ) g_g 2_g 
and some phosplioric acid . . . . ) 
Carbonate of lime 12-8 1-0 4-G 
Sidphate of lime 2-i-l 0-54 3-8 
Alkaline sulphate, wllli trace of chloride) „ , i i> o 
1,1. f U • 1 I' Ml U • o 
anu suiplunct I 
Water 8-8 4-G 4-(i 
Trace of magnesia, and loss .... 1'7G 1*4 2-4 
lUU.O l(li)-l) 100.0 
