688 
British Association. 
[October, 
posed Compound Nature of the Elements.” Continuing his 
researches on this subject, he found that when carefully 
distilled metallic sodium is condensed in a capillary tube, 
placed in a retort, and heated in a Sprengel vacuum, it 
gives off twenty times its volume of hydrogen. Phosphorus, 
carefully dried, gives off 70 volumes of gas, chiefly hy- 
drogen. A specimen of magnesium, carefully purified by 
Messrs. Johnson and Matthey, was magnificent in its colour- 
ings ; it gave first hydrogen, then the D line, then the green 
lines of magnesium (6), then the blue line, and then, as the 
temperature was increased, various mixtures of all of them, 
the D line being always the most brilliant. In this experi- 
ment only two volumes of hydrogen were collected. From 
gallium and arsenic no gas is given off. From sulphur and 
some of its compounds sulphurous anhydride was always 
obtained. From indium, hydrogen was given off before 
heating ; while from lithium no less than 100 volumes of 
hydrogen were given off. 
Dr. Gilbert, F.R.S., read a paper “ On some Points in 
connection with Agricultural Chemistry,” in which he stated 
that Mr. Lawes, at Rothamsted, and himself, had now 
grown wheat for 36 years in succession on the same land, 
barley for 28 years in succession, oats for 10 years, root 
crops for more than 30 years, beans for more than 30 years, 
and they had experimented with mixed herbage on grass 
land for 24 years. They found only minor distinctions in 
the manurial requirements of different plants of the same 
natural family, but very great distinctions in the require- 
ments of plants of different natural families. The gramineous 
crops are very low in the percentage of nitrogen, and yield 
but a very small quantity per acre. Yet nitrogenous ma- 
nures are very effective when applied to such crops. 
Leguminous crops, on the other hand, are very high in the 
percentage of nitrogen, and yield a very large amount of it 
per acre. Yet nitrogenous manures are of little avail to 
these plants, and potash manure is especially effective. 
The differences in manure requirements of plants of all 
natural families were also pointed out. Much more com- 
plicated, however, was the problem when experiments were 
made upon mixed herbage of grass land, where they might 
have fifty or more species growing in association, repre- 
senting perhaps twenty natural families. One point of 
especial interest was the difference in the amount of nitro- 
gen yielded by the plants of the different natural families. 
It was assumed by some that some plants assimilated the 
free nitrogen of the atjnosphere ; but so far as existing evi- 
