Tubes were also prepared containing grass immersed in a 1 per cent. solution of phenol, and a 1 per cent. solution of mercuric chloride respectively, but in neither case was any gas evolved. A tube containing grass immersed in water was exposed to the heat of a steam-bath for several hours, and this yielded no gas, although the tube was kept for several weeks. Thus the evolution of gas is evidently dependent upon the presence of low forms of organic life, and in the water taken from the tubes, the microscope always revealed numerous bacteria. As this water also gave an acid reaction with test-paper, a further experiment was made on a larger scale to determine the other products of the fermentation.

For this purpose about 7 lbs. of grass were introduced into a capacious retort and covered with recently boiled water. During the first three days the contents of the retort were exhausted with a waterpump, after which the retort was fitted with a delivery-tube, and the gas subsequently evolved was collected over water. In the course of 22 days about 1 litre of gas had collected, which on analysis was found to contain 20 per cent. of hydrogen.

After 27 days the liquid was poured off and distilled with phosphoric acid to a small bulk. The distillate, containing the volatile acids, was treated as below. The residue remaining in the retort was filtered, and the filtrate repeatedly shaken out with ether in order to extract the fixed organic acids. The ethereal extract was then diluted with water, an excess of precipitated baric carbonate added, and the liquid heated to volatilise the ether. After filtering off the excess of baric carbonate, the filtrate was evaporated to dryness on a waterbath, and the residue, after being washed with alcohol, was dried at 130° C., and the barium determined:

0-2255 gram of barium salt gave 0.1655 gram BaSO1, equivalent to 43.14 per cent. Ba.

Baric lactate.

(C,H,O3)2Ba.

Ba per cent.
calculated.

43.49

Found.
43.14

The distillate mentioned above, and containing the volatile acids, was divided into two equal parts, one of which was then exactly neutralised with caustic potash. After again uniting the two portions,

VOL. XLIII.

Y

the mixture was distilled nearly to dryness. The acid distillate was treated with an excess of freshly precipitated baric carbonate, and, after filtration, the solution was evaporated down, and the residue recrystallised. The crystals were dried at 130° C., and the bariumdetermined :

I. 0.4540 gram of barium salt gave 0.3757 gram BaSO,

per cent. Ba.

= 48.65.

II. 0-6080 gram of barium salt gave 0.5028 gram BaSO, = 48.61

Although the barium salt thus agrees in its percentage of barium with baric propionate, there is just the possibility that it might have been a mixture, in the proportion of their molecular weights, of baric acetate and baric butyrate; unfortunately sufficient material was not at our disposal to decide this point.

The residue left in the retort in the last distillation was treated with phosphoric acid and distilled almost to dryness. The acid distillate was treated with an excess of baric carbonate, filtered, the filtrate evaporated down, and the residue obtained recrystallised. The crystals were dried at 130° C., and the barium determined :

0.2950 gram of barium salt gave 0.2674 gram BaSO1 = 53-28 per cent. Ba.

Baric acetate. (C2H2O1),Ba.

Ba per cent.
calculated.

Found.

53.73

53.28

The products of fermentation which accompany the evolution of gas described are thus acetic, lactic, and probably propionic acids. From these experiments it appears:

(1.) That comparatively dry grass soon evolves considerable quantities of gas consisting almost wholly of carbonic anhydride, and accompanied by mere traces of hydrogen and hydrocarbons.

(2.) That the evolution of gas takes place with almost equal rapidity in atmospheres composed of air, carbonic anhydride, oxygen or hydrogen; and that excepting when the atmosphere contains oxygen, in which case a notable proportion of nitrogen accompanies the carbonic anhydride, the composition of the gas evolved is much the same.

(3.) That when the decomposition of grass takes place under water,

large volumes of gas are evolved which are characterised by the presence of a notable proportion of hydrogen. This hydrogen is doubtless due to the lactic fermentation induced by bacteria; acetic, lactic, and probably propionic acids, together with bacteria, being found in the water in which the grass was immersed.

XXXVII.-Note on an Apparatus for Fractional Distillation under Reduced Pressures.

By L. T. THORNE, Ph.D.

IN distilling in vacuo, or under reduced pressure, the chief difficulty is usually the removal of the various fractions without breaking the continuity of the distillation. It is believed that the apparatus of which a sketch is annexed, will be found to be an improvement on those now in use.

The receptacle e is a tube about 10-12 cm. long, and 10-12 mm. in diameter, closed below with a stopcock c, the lower delivery tube of

which is drawn out to a narrow quill tube about 4 cm. long.* Near the top of e is a tubulure d to admit the end of the condenser; the upper end is somewhat narrowed for convenience in making connection with the stopcock a. The other end of the stopcock a is connected with a T-piece h, the outlet i of which communicates with the exhausting pump, and the third limb with the three-way cock b. To the other end of b is attached the tube k, the lower end of which is slightly drawn out and bent parallel with the quill end of c. These two tubes are then passed through a doubly bored caoutchouc stopper f, on which is fastened the bottle, test tube, or other receiver m.

The drawn-out end of the condenser or cooling tube n is fitted airtight into the tubulure d by means of a cork, the end of the tube projecting well into e: the outlet i is connected with the vacuum pump, stopcock a being open, and the three-way tap b so arranged that h and k are in communication, and the pump is set in action. In this way the whole apparatus is exhausted, and the distillation is then commenced. As soon as the first fraction has passed over into e, or that receptacle has become full, stopcock c is opened, when the distillate at once flows into m. Should the distillate be at all viscid and not flow easily, stopcock a is closed, and the exhauster acting through b and c assists gravitation: c is then closed again, and if it is wished to change the receiver m, the three-way tap is so turned that m is put in communication with the air while the bore leading to his closed. When a fresh receiver m' has been adjusted, a is closed for a few seconds, and b turned so as to place h and k again in communication, and when m is exhausted a re-opened. In this way any number of fractions may be collected without for a moment stopping the distillation, and by selecting ƒ of a suitable size, the distillate may, if desirable, be collected directly into the vessel in which it is afterwards to be used.

In practice it is found that owing to the slight irregularity in working of the exhausting pump (especially where a water-pump is used), it is very advantageous to insert between i and the pump a reservoir of two or three litres capacity. This reduces the effect of these irregularities very much, and if this reservoir be placed in direct communication with a mercury pressure-gauge, any desired reduction of pressure can, with a little care, be obtained and maintained constant. If the substance under distillation is affected by exposure to the air, any inert gas may easily be introduced instead of air by means of the three-way tap when changing the receiver.

This has been found a convenient size for general laboratory use, but can of course be modified to suit special circumstances.

303

XXXVIII.-Notes on the Condition in which Carbon exists in Steel.

By Sir FREDERICK ABEL, C.B., F.R.S., and W. H. DEERING, F.C.S.

THE experiments, the results of which are given in this paper, were made for the Committee on Steel of the Institution of Mechanical Engineers, in the hope of obtaining information on the condition in which carbon exists in steel as it is left by the cold-rolling, and in its hardened, annealed, and intermediate conditions.

The results obtained are considered as preliminary. The earlier series of experiments having showed differences in the behaviour of the hardened steel as compared with the cold-rolled and annealed steel, and in the amount of carbide of iron left by the oxidising solution used, the second series of experiments was devoted to ascertaining the limits of strength of the chromic solution within which approximately the same percentage of carbide of iron would be obtained.

First Series of Experiments.

The steel was used in the form of discs 2.5 inches in diameter, and 0.01 inch thick. Twelve discs were supplied by Mr. Paget of Loughborough, all of which had been cut out of the same strip of metal, the odd-numbered discs (1, 3, 5, &c.) having been cut from one side of the axis of the strip, the even-numbered discs (2, 4, &c.) from the other side. The weight of each disc was about 6.5 grams.

The discs numbered 1, 4, 7 and 10 were the steel as received from cold-rolling. Nos. 2, 5, 8 and 11 were annealed. Nos. 3, 6, 9 and 12 were hardened.

Mr. Paget describes the hardening process as follows. The discs to be hardened were placed between two cast-iron blocks, one being recessed to receive the plates, and the other being quite flat. These blocks were equally heated to a bright red; a disc was then placed between them, and allowed to remain there until thoroughly and equally heated; it was then instantaneously removed, and as quickly as possible caught and pressed between two cast-iron surface-plates.

In the annealing process, the discs were bolted between wroughtiron plates 3 inch thick, and these were then enclosed in a thin sheetiron box (5 inches square by 2 inches deep). This sheet-iron box was placed in the centre of a cast-iron box (about 15 inches x 6 inches, and about inch thick). The space between the two boxes was filled up with flue-dust (i.e., thoroughly burnt soot taken from the flues of a boiler near to the fire end). The whole apparatus was then raised in an annealing furnace to a bright-red heat, sufficient to scale the cast