H2SO4. HCl. Sulphate. Chloride. Sulphate. Chloride.

H2SO4 HCl. Sulphate. Chloride. Sulphate. Chloride.

The above results show that although the absolute amounts of absorption are different, the ratios of absorption are the same for wool and silk.

On account of the very small absorption by cotton of the individual acids (see Tables VIII and IX), we did not submit that material to the action of their mixtures.

The mean probable error of a single comparison with experiment of our calculated nine equations is 0.00075 gram, or about 2 per cent. on the mean total absorption. As this is about the ordinary error of careful work with small quantities, we infer that the absorptions we have measured are amenable to laws already established in other fields of chemical investigation. Combination proceeds at first with considerable absolute rapidity, and continues with decreasing rapidity. Increase in the mass of reagent—which as a rule accelerates combination-may be counterbalanced by lowering the temperature (Tables IVa and IVb).

From the column "maximum absorption per gram" in Table XI, it is easy to calculate that the ratios of such absorption in the last six (comparable) experiments are―

numbers exhibiting a significant tendency to identity in the case of all three reagents.

Lines 5 and 8 in the same table show that wool and silk also tend to resemble each other in the weight they absorb of sodic hydrate; but

from the remainder of the table it is clear that wool takes up much more from acid solutions than is the case with silk.

The quantities 0.41810 gram hydric chloride and 0.45925 gram sodic hydrate in the same table are in the ratio HCl : NaHO. When wool is treated with the weak reagents severally in this proportion, the absorption is nearly in the ratio 2HC1: 3NaHO. The corresponding results for silk and cotton are 3HC1: 10NaHO in both cases. There is thus a very intimate relation between silk and cotton-a relation which, whatever it may be in part, is shown by these changes to be, to a great extent, of a strictly chemical nature.

We have, in conclusion, to express the hope that our investigation, while bearing on the one hand on questions of great technical importance, may not be without its value in the profounder future study of cotton, silk, and wool-three bodies of definite chemical composition, but whose intimate constitution still remains obscure.

XXI.-On the Action of Chlorine on Certain Metals.

By RICHARD COWPER, A.R.S.M., Demonstrator in the Laboratory of the Royal Naval College.

IN performing the ordinary experiment of burning sodium in chlorine, I observed that with a more than usually dry sample of the gas, the bright surface of the melted metal tarnished very slowly; and it occurred to me that if the chlorine were in a perfectly dry condition it might have no action, and also that this might be true in the case of other metals said to combine directly with chlorine at ordinary temperatures.

I have since found that this inability of dry chlorine to act on sodium has been observed by Wanklyn (Chem. News, 20, 271), who has not, however, so far as I can discover, pursued the experiment with other metals.

I proceeded as follows:

Chlorine was prepared by the action of pure hydrochloric acid on manganese dioxide. The gas was passed through three wash-bottles containing water to free it from hydric chloride, and then through tubes of from 4-inch to 1-inch diameter, having an aggregate length of about 8 feet, closely packed with anhydrous porous calcium chloride.

It is well known that if Dutch metal be introduced into a vessel containing chlorine prepared in the usual manner, combustion takes place with evolution of light and heat. It was found that chlorine

prepared and dried in the manner above described acted very slowly on Dutch metal.

Some Dutch metal was next placed with a piece of freshly fused calcium chloride in a small thin glass tube and sealed. This small tube was placed in a much larger tube, also containing some pieces of freshly fused calcium chloride. The larger tube was then filled with chlorine dried by passage through calcium chloride tubes, then sealed, and allowed to stand for several days. The small tube was then broken by shaking, and the metal thus exposed to the chlorine. It was found that the Dutch metal remained bright. (A specimen thus prepared is still apparently unacted on after about three months.)

On introducing a minute quantity of water into the tube, the chlorine was rapidly absorbed. If a drop of water were allowed to come in contact with the metal, the action was instantaneous and accompanied by evolution of light and heat.

When chlorine, dried by passage through calcium chloride tubes, was brought in contact with dry metallic zinc in the form of thin foil, the latter was distinctly acted on, assuming a moist appearance; but this was not the case when the chlorine had stood for some days in contact with freshly fused calcium chloride.

It was observed that the zinc was most acted on by the partially dried chlorine at the point where the gas first came in contact with the metal. A glass tube about 18 inches in length was closely packed with pieces of thin zinc-foil, and chlorine dried by calcium chloride was passed through it. It was found that the action, as indicated by the moistening of the surface of the metal, extended but a very short distance from the end at which the gas was admitted, the greater part of the zinc remaining dry and bright. Metallic zinc, therefore, appears to offer a highly advantageous means of removing the last traces of moisture from chlorine.

Experiments were made with the following metals, the chlorine in each case being allowed to stand in contact with pieces of freshly fused calcium chloride in a sealed tube during several days, in the manner already described:

Magnesium, in the form of wire, was not attacked.

Silver, in the form of leaf, was acted on very slowly. The presence or absence of light appeared to make no difference in the rate at which the metal was attacked.

Bismuth was at first apparently unacted on. After several days, a specimen in the form of coarse powder had become slightly tarnished, but the atmosphere in the tube remained quite yellow.

Tin, in the form of foil, was rapidly attacked with evolution of heat. Antimony and arsenic, in the state of powder, were acted on immediately, with evolution of light and heat.

It may be noted that these last three elements produce chlorides which are liquid at ordinary temperatures.

Mercury appeared to be acted on by the dry chlorine as rapidly as by the moist gas.

It has been observed by Wanklyn, as mentioned above, that sodium is not acted on by dry chlorine, either at ordinary temperatures or when melted in contact with the gas.

I made the following experiment:-A piece of sodium was placed in a glass tube. Chlorine, dried over calcium chloride, was passed through the tube, and the piece of sodium was heated to dull redness. It was then allowed to cool in the current of chlorine, and the tube was sealed. The sodium was then raised to the melting point, and shaken so as to cause the metal to flow from its slag-like envelope. At first it became dulled; but after several times causing a fresh surface to be exposed to the chlorine, it remained bright, the atmosphere in the tube being at the same time quite yellow, showing that very little chlorine had been absorbed. The action observed on first exposing a bright surface of melted sodium to the chlorine is no doubt due to the gas being imperfectly dried.

If dried chlorine be passed over a piece of potassium, the latter catches fire. This is, however, probably due to the heat caused by the action of the chlorine on the envelope of KHO, which surrounds the piece of potassium.

Some potassium was next sealed up in a tube containing dry air. The metal was then heated until all the oxygen was absorbed, and a bright surface of potassium obtained. The tube was then filled with chlorine, dried by passage first through calcium chloride tubes, and then through a tube packed with thin zinc-foil. The potassium at first remained bright, but slowly became covered by a film of a rich purple colour. This is no doubt the sub-chloride described by Rose

(Pogg. Ann., 121).

The action was much accelerated by raising the metal to its melting point. The potassium did not, however, burn until the temperature greatly exceeded the point of fusion.

If the purple compound be heated in vacuo, it is decomposed, metallic potassium and potassium chloride being apparently formed. If a minute quantity of water be introduced into a tube containing potassium and chlorine, the latter is rapidly absorbed.