044 MR. J. B. LAWES, DR. GILBERT, AND DR. PUGH ON 
previously covered portions of the slate or stone-ware lute were then washed with pure 
distilled water, and the wash-water was added to the condensed or drain-water. In the 
experiments of 1858 this fluid was analysed separately, but in those of 1807 it was 
mixed and dried down with the soil. 
The pot, with the soil and plants, was removed to a clean table covered with white 
paper, the plants measured in all their parts and then cut off at the surface of the soil; 
the roots were removed, slightly washed from soil, and observed. The plants were then 
put into a small wide-mouthed bottle, generally stem and root together, but sometimes 
they were put into separate bottles. In the experiments of 1857 the contents of the 
bottles were dried in a water-bath, with a current of air, previously washed through 
sulphuric acid, passing through the bottle and thence through a solution of a known 
quantity of pure oxalic acid. But it was found that no appreciable amount of ammonia 
was thus accumulated. Hence, in 1858, a little oxalic acid (in solution) was added to 
the vegetable matter, and the whole dried in the water-bath without the above pre- 
caution. 
When dry, the vegetable matter was cut small by means of a pair of clean long 
scissors, reaching to the bottom of the bottle. In this way the substance was reduced 
to a considerable degree of fineness, and it was still further ground up in the mortar 
when mixed with soda-lime for analysis. When duplicate analyses were to be made, the 
matter was carefully divided so as to ensure equal proportions of stem, fine leafy matter, 
&c., in each half. Hence, if both analyses were successfully conducted, the results were 
mutually confirmatory; or if one portion were lost, the other still represented a propor- 
tionate amount of the whole material. 
The soi? was removed from the pot to a porcelain dish, and a sufficient amount of a 
solution of oxalic acid added to keep it acid. The mixture was then heated on a sand- 
bath (stirring constantly) until most of the water was expelled, more fully dried in a 
water-bath, and then preserved in well-corked bottles for analysis. The pots were pounded 
up; those of 1857 being preserved and analysed separately, and those of 1858 mixed 
with the soil before it was dried with oxalic acid. The pieces of flint at the bottom of 
the pot were also pounded and mixed with the soil. 
For analysis, 150 to 200 grammes of the soil, pot, or mixture, were mixed with about 
half the volume of soda-lime, the whole put into a large combustion-tube, some soda- 
lime put in advance of the mixture, and then asbestos, as usual. The combustions 
were made in charcoal furnaces, and the ammonia collected in titrated sulphuric acid, 
of which the strength, and the amounts employed, are described at pp. 545,546. When 
very small quantities of nitrogen were involved, the ammonia from two or three tubes of 
substance was sometimes collected in the same quantity of acid, so as to diminish the 
error of titration. It was found, however, to be better to use very small quantities of 
acid, and to estimate the product of each combustion separately; for, by the former 
method, if any accident occurred in the second or third combustion, it involved the loss 
of the determination of the products previously collected. 
