From the above table, it will be seen that Dyer's method shows twice as much available potash in the Group I soils as in the Group II, and this result agrees well with the results of the manurial experiments, which showed so clearly that Group I soils contained much more available potash than Group II.

The agreement would have been still more striking had not the occupier of the land at Flitcham applied two dressings of chloride of potassium, between the times when the field experiments were made there and the samples of soil were taken for analysis.

Dyer's method thus shows clearly the very different amounts of available potash in the two groups of soils; indeed, had it been applied to the Flitcham soils in the first instance, it would have indicated, more rapidly, and more economically than the field experiments did, the necessity of applying soluble potash manures in order to grow a profitable crop.

At the end of Dr. Dyer's paper, there is a suggestion that, in order to make the results strictly comparable for all soils, sufficient citric acid should be added to neutralise all the chalk in the soil and leave 1 per cent. over.

As several of the above soils contained large amounts of chalk, I acted on this suggestion, but found that, under these experimentat conditions, the amount of potash dissolved was nearly equal in the case of each of the three soils tried. The numbers are annexed.

In adding the extra citric acid, the following method of procedure was adopted. The soil was put into a Winchester quart, with 1 per cent. acid, as detailed in Dr. Dyer's paper, and the amount of citric acid required to neutralise the chalk was weighed out and added at the rate of about 1 gram per hour, with frequent shaking.

When most of the extra acid had been added, the solution was titrated, and found to be nearly 1 per cent. in strength. It was then left until the sixth day and titrated again, and a further small amount of acid added to make the strength 1 per cent. On the seventh day, the liquid was separated, and the dissolved potash determined as before.

As will be seen by the above numbers, this method gave no indication of the much greater amount of available potash in the

Higham soil than in the two soils from Flitcham, and I venture to suggest that the clearest indication of the amount of available potash is obtained by extracting the soil with 1 per cent. citric acid, without regard to the amount of chalk contained in it.

It seems to me, also, that by doing so one is imitating more nearly the conditions under which the plant obtains its potash, for there is no evidence to show that the acidity of the root juice is greater when the soil is more calcareous.

Available Phosphoric acid.

Among the results of the Norfolk and Suffolk experiments are many numbers referring to the growth of swede turnips with different manures, and I have picked out the following as giving a graduated series in which the effect of superphosphate is practically nil, very great, and intermediate between these two.

It will be seen that the effect of superphosphate is practically nil at Bramford, while at Higham it is considerable, and at Warham greater still.

I have estimated the phosphoric acid in these three soils by Dyer's method, both with and without the addition of extra citric acid to neutralise the chalk, the extra acid being added in the manner described above. The results are given in the following table.

I being phosphoric acid soluble in strong nitric acid on boiling;

II, phosphoric acid soluble in 1 per cent. citric acid; III, phosphoric acid soluble in 1 per cent. citric acid using extra acid to neutralise the chalk.

The analytical numbers obtained with the 1 per cent. acid show very clearly the much greater amount of available phosphoric acid in the Bramford soil than in either of the others, and the much smaller and more nearly equal amounts in the soils from Higham and Warham. This agrees perfectly with the results obtained in the field experiments with swedes quoted above, and the method appears to give even better results for phosphoric acid than it did with potash.

When the extra citric acid was added, in the case of the Bramford soil with only 1.97 per cent. of lime, a slightly greater amount of phosphoric acid was dissolved; in the Higham soil, containing 5.88 per cent. lime, nearly five times as much P2O, was dissolved as with 1 per cent. acid only. The Warham soil contained so little lime that no determination with extra acid was thought necessary.

Comparing the results obtained with and without the extra acid, it appears that the 1 per cent. acid only, the chalk being neglected, gives numbers more nearly proportional to those ascertained by field experiment; and again I think that this is only to be expected, for thus we imitate most nearly the solvent action of the plant juices.

In the case of the 1 per cent. acid and a soil rich in chalk, the large amount of carbonic acid evolved must have an appreciable solvent action on the potash and phosphates present. 100 grams of the Higham soil suspended in 1 litre of water through which purified carbonic anhydride was passed for 48 hours yielded the following numbers.

K2O dissolved by water saturated with CO2....
P2O

K2O soluble in 1 per cent. citric acid

....

0.008 per cent.

0.011

....

0.013

My thanks are due to Dr. T. H. Easterfield for many valuable suggestions given in the course of the work.

Agricultural Department,

University Chemical Laboratory,
Cambridge.

293

XXX.-The Production of Naphthalene and of Isoquinoline Derivatives from Dehydracetic acid.

By J. NORMAN COLLIE, Ph.D., F.R.S.E., and N. T. M. WILSMORE, M.Sc. (Melbourne).

In a former paper by one of us (Trans., 1893, 63, 329 et seq.), it was shown that under certain conditions diacetylacetone condenses with loss of water, forming a yellow, crystalline compound, melting at 108-109°, and this compound condenses further to a second yellow substance, which melts at 183-184°, and proved to be a naphthalene. derivative. All the work then done with this second substance indicated, with considerable probability, that its constitution was 3: 3'-dimethyl-2-acetyl-1: 1'-dihydroxynaphthalene.

.

The present communication is an account of further research on these compounds, in which it was sought to elucidate the mechanism of the reactions taking place in their formation, and to obtain further evidence as to their constitution.

It would appear from what follows that, of the possible tautomeric forms of diacetylacetone, one is comparatively stable and incapable of condensation to the above compound, whilst another is unstable— readily changing into the stable form, or condensing by union of 2 mols. with loss of water. It is interesting to note in this connection that W. H. Perkin (Trans., 1892, 61, 827) has shown, from the magnetic rotation, that at a temperature of 60.4°, two double linkings are present in the diacetylacetone molecule, whilst at lower temperatures, it tends to pass into the trihydroxy-derivative.

The condensation occurs most readily when the diacetylacetone-which acts as a dibasic acid to strong bases-is combined with only half its equivalent of barium. Possibly 2 mols. are here united by an atom of barium in the form of an acid salt, this being a first step to the more complete union. Certainly a univalent metal, such as sodium or potassium, cannot be substituted for barium; but, on the other hand, the formation of the yellow substance will take place, even if the whole of the barium be removed as carbonate by a current of carbonic anhydride. In all other cases, the barium hydroxide liberated by the condensation hydrolyses some of the remaining diacetylacetone, forming barium acetate. The condensation is, however, not complete in any case; and at most, only half of the theoretical yield has been obtained.

In accordance with the foregoing, we would suggest that the normal barium salt of diacetylacetone should also be written double,

VOL. LXIX.

Σ

Ba,

-O-C(CH,):CH-CO-CH:C (CH,)O, having the constitution Ba<O.C(CH3):CH.CO·CH:C (CH2)0instead of that suggested by Feist (Annalen, 1890, 257, 276). Owing to the difficulty with which the yellow compounds react, comparatively little progress has been made with them. From the second, dimethylacetyldihydroxynaphthalene-a dimethylnaphtha. lene has been prepared, which, on oxidation, yields 1: 3: 4-methylphthalic acid. Also, by the action of strong sulphuric acid, the acetyl group appears to have been removed, leaving a colourless substance.

Very little direct experimental evidence as to the constitution of the first substance, m. p. 108-109°, has been obtained. In all probability, it is a benzenoïd compound, having long side-chains which unite to form dimethylacetyldihydroxynaphthalene. The condensation of diacetylacetone, therefore, takes place in two stages.

The probability of the benzenoïd character of the first compound is heightened by its peculiar behaviour with ammonia, resulting in the formation of what appears to be a derivative of isoquinoline. The reaction seems to be

By heating this new base with strong sulphuric acid, part of the side-chain is removed, leaving aa'-3-trimethyl-1-hydroxyisoquinoline (1' 3' 3-trimethyl-1-bydroxyisoquinoline),