cools should remain in contact with the products of condensation, which are being cooled at the same time, and which are exactly those intended to absorb the ammonia. With the reverse motion the cooling action would, no doubt, be increased; but the scrubbing effect would be null, since the products of condensation would be heated by the ascending gas, and would part again with their ammonia. The wood shavings serve to attract the vapours mixed with the gas, which vapours when deposited over an extended surface further assist in arresting the ammonia. Both gas and ammoniacal water leave the scrubber at the bottom.

In large works where machinery and superintendence are comparatively small items in the balance sheet, and ammonia is valuable, this steam jet scrubber will not offer great advantages. It is quite different, however, in small and medium-sized works. There the trifling loss of ammonia is of no consequence, whereas the freedom from machinery and skilled labour exerts a most beneficial effect upon the economical production of gas; and for such works, therefore, the present scrubber is to be recommended.

The gas on entering the scrubber ought to be as free from tar as possible; for if through imperfect condensation a large amount of it should still remain in the gas, the rapid motion and simultaneous heating produced by the steam jet would decompose the minute particles of tar, the volatile oils would be retained in the gas in a state of solution, pitch and naphthalin would be formed, and choke up the scrubber in a short time. If the gas is free from tar the wood shavings may remain in the pipes for years without renewal. In Liverpool, for instance, the same shavings have remained for three years in one scrubber; but the system of condensation adopted there separates the tar and precipitates the ammonia thoroughly.

The first point of importance in the scrubber is the equal flow of the gas through all the pipes at once, as also of the water coming from the exhauster, and of what is separated from the gas in the main horizontal pipe. To effect this the vertical pipes are narrowed at the top by a disk carrying a smaller pipe, projecting into the horizontal main by about a quarter of its diameter. The areas of these small pipes together should be equal to that of the main in the works. The second point of importance is, that the liquid particles deposited by the gas in its downward flow should be directed on the middle of the shavings, which is accomplished by inserting a disk with a hole in the middle at the joint of two pipes.

The area required for the scrubbers to absorb the ammonia sufficiently depends upon climate and locality; but, on an average, a production of 1,000 feet per hour requires from 100 to 150 square feet of pipe surface.

The cost of these scrubbers, as compared with those of Man and Walker, is about the same for large works, and for small ones about two-thirds; and their great advantage consists in the absence of all machinery, with the exception of the boiler.

Tests made at Berlin as to the amount of ammonia contained in the gas after it had left the scrubber, and after passing it through a refrigerator in the shape of a scrubber with a finely divided spray of water, showed an average of 23 grammes of ammonia in 3,530 feet of gas. When, instead of this second scrubber, a pipe condenser surrounded by water was substituted, the result was an average of 51 grammes of ammonia in the same quantity of gas. This is a proof that a spray of water is a most valuable contrivance for removing the ammonia. Experiments made at the gasworks at Dessau corroborate the above, it having been found that gas, after passing the first scrubber, lost 5 per cent. of its ammonia, after the second it lost 16 per cent., after the third 33, and after having gone through a fourth scrubber it had lost 84 per cent. The dimensions of the scrubbers with which these experiments were instituted were 8 feet in diameter by 20 feet in height, with plates inserted at intervals covered with twigs.

J. G. H

On the Manufacture of Gas from Paraffin Oils.

By L. GROTOWSKY.

(Zeitschrift für das Berg-, Hütten- und Salinen-Wesen, vol. xxv., pp. 176–182, 1 pl.)

The raw material employed in the production of illuminating gas from liquid hydrocarbons is the residual oil from the manufacture of paraffin, known in the trade in England and Galicia, according to the character of its fluorescence, as blue or green oil, and in Germany as paraffin oil. This, when allowed to flow over a metal surface heated to dull redness, is decomposed and volatilised, the products being partly condensable, mainly tarry vapours, and partly permanent gases of a very high illuminating power. The proportion of tar to gas is about 1 to 2, or a given weight of oil will produce from 60 to 70 per cent. of gas.

The distillation is effected in retorts, which may either be horizontal and generally similar to those used in making coal, or vertical like those employed in the distillation of lignite oil shales, &c. In the former or horizontal construction, the retort is a castiron tube of oblong section, about 5 feet long, 10 inches broad, and 5 inches high, supported along its entire length by a bench of firebrick, and heated by a grate fire placed below, the sides and top being surrounded by the flame, but not the bottom. The oil is introduced at the hinder end through a tube of the bore of a straw or quill, laid with a slight inclination upwards, that the stream may strike the roof and flow down the sides close to the point of admission. The flow is regulated by a cistern, in which a constant head of oil is maintained. The volatile products are led off by a still-head and goose-neck above the fireplace end to a box containing tar, where a first separation of the tarry vapours

ON THE MANUFACTURE OF GAS FROM PARAFFIN OILS. 389

is effected, and subsequently to the condensers proper (which are vertical cylinders filled with fragments of coke), where the rest of the tar is condensed; after which the gas is sufficiently purified for use, and is received in a gasholder of the ordinary construction.

The vertical retort is a cast-iron tube, slightly reduced in diameter at the lower end, set in a furnace with a system of ring flues, so that it may be uniformly heated. The fireplace is below, and at one side of the axis of the retort. A wrought-iron cylinder, similar to that employed in charcoal blast furnaces for the same purpose, serves as a collector for the gas. This is suspended from the top flange of the retort, and is removable, so that it may be taken out and cleaned when necessary. The oil is introduced through a series of small-bore tubes, penetrating the wall of the retort through the top end, and disposed at equal distances around the circumference. The gas collecting and purifying arrangements are generally similar to those in the horizontal form of apparatus.

The vertical construction is considered to be preferable to the horizontal, as giving a larger yield from the same sized plant; an apparatus of the dimensions shown in the accompanying plate with the horizontal retort, producing 600 cubic feet per hour, while that of the vertical form is 750 cubic feet, from oils equivalent to 1,000 cubic feet. In the latter construction it was considered desirable to case the retort with firebrick, either wholly when it is kept in constant use, or partially when it is used intermittently.

The chief care required in working is the maintenance of a proper temperature in the retort, which should be uniformly of a dull red (brick red) heat. If it is irregularly heated, besides the danger of breakage there is a proportionate increase in non-illuminating hydrocarbons and tar, so that the yield is depreciated both in quality and quantity. If uniformly overheated, the decomposition is attended with a separation of carbon, which is deposited as soot in the connecting tubes. When, on the other hand, the heat is too low, there is, in addition to an excessive production of tar, a formation of asphalt, which deposits in the retort itself. The most favourable results are obtained with a yield of 900 or 1,000 cubic feet of gas per cwt. of oil, as, although a more highly illuminating gas may be produced when the yield is smaller, the total illuminating factor per unit of oil consumed will be less; and conversely, with a high quantitative yield, the illuminating value of the gas will be diminished. The firing is regulated chiefly by the colour of the gas as observed through gauge cocks in the still-head, which should be of a clear brown tint. If it is white the temperature is too low, and if colourless with spots of soot, it is too high. The production of tar should be from 26 to 30, or not exceeding 33 per cent., and the tar condensers should keep tolerably cool.

The gas produced is of a much higher density and illuminating power than ordinary coal gas, being very rich in elayl and similar

light-giving hydrocarbons. The following are the comparative figures of various kinds of gas used for lighting:

The illuminating power of the flame from a standard burner of 1 cubic foot consumption of oil gas per hour is equal to that of thirteen standard candles of 100 grammes; the corresponding value for coal gas being 2 44, or the illuminating power of the former is 5.33 times that of the latter; or for equal cost of lighting, oil gas at 218. 8d. per 1,000 feet would be as cheap as coal gas at 5s. 6d. per 1,000 feet.

At the factory of Gerstewitz, near Weissenfels, oil gas is used to the extent of 880,000 cubic feet per annum. The gasification is attended with a production of 26 to 33 per cent. of tar, which by rectification yields 1.33 per cent. of benzole, 1.9 of chemically pure naphthaline, and 65 per cent. of condensed products suitable for conversion into gas at the rate of 780 cubic feet per cwt. The loss of gas in the pipes does not exceed 1 per cent., owing to its high density and the low pressure required at the burners.

A single retort, with a gasholder of 375 cubic feet capacity, suffices for the supply of thirty to seventy burners, or up to one hundred with the gasometer volume increased one-half. Two retorts, with 1100 cubic feet of gasometer space, supply one hundred to two hundred burners. The cost of the entire apparatus in the first case is £65; in the second, £90; and in the third, £137 10s. (the mark taken at one shilling).

Oil gas plant of the above description has been supplied to thirty-five establishments, principally in France, with a total of six thousand burners, by C. W. Schumann, of Weissenfels; and to seventy-four (up to July 1874) by P. Suckow and Co., of Breslau, the latter including gasworks for six towns, fifty-eight factories, and one lignite mine.

H. B.

On the Magnetic Constants of Nickel. By M. WILD. (Bulletin de l'Académie des Sciences de S. Pétersbourg, vol. xxiv., pp. 1–35.)

The following investigation was made upon a nickel magnet, prepared by Joseph Wharton, of Philadelphia, and presented to the Author by P. Kotschubey, President of the Russian Technical

Society. It is in the form of a flat bar with pointed ends (2 millimètres thick, 9.5 millimètres broad, and 155 millimètres long), with an agate cap in the centre for suspension on a pivot. The weight is 25 grammes without the cap.

The knowledge of the magnetic properties of this metal is somewhat indefinite. According to Biot the permanent magnetism of a nickel needle is only one-third of that of one of steel of similar size. Lampadius determined the attractive force of a magnet on equalsized masses of nickel and iron to be as 35 to 55, and Gay-Lussac considered the momenta of induction by a magnet on nickel and iron to be to each other as 1 to 2. Lastly, Arndtsen found that the momenta of induced and temporary magnetism in nickel are not susceptible of increase by augmentation of intensity in the magnetising power, beyond a maximum which is soon reached, and that the former is to the latter as 205 to 94.

In the present series of experiments the comparisons were made with a steel magnet, belonging to a unifilar magnetometer by Krause, which was nearly of the same weight (30 grammes), and of about the same dimensions as the nickel magnet; and subsequently with three others of Wolfram steel; the angle of deviation of a compass needle by the deflecting influence of the different magnets acting at right angles to the magnetic meridian at a uniform distance of 300 millimètres, forming the standard of comparison. By analysis the nickel was found nearly but not absolutely pure; it was free from copper, manganese, and cobalt, but contained 0.33 per cent. of iron. The experiments included determinations of the specific magnetic momenta of the bar, under the following conditions: 1, as received; 2, immediately after remagnetisation; 3, two days after remagnetising; 4, after two months and a half; and, 5, after alternate heating and cooling-the latter being for the purpose of determining the temperature co-efficient. The relation of the permanent to the temporary magnetism was obtained by comparing the deflection produced by the nickel bar when employed at the core of an electro-magnetic helix, which was alternately put in and out of the circuit of a battery of known intensity; another term of comparison being obtained by the substitution of a soft iron bar of the same weight under similar conditions.

=

The absolute magnetic momentum of any particular magnet is found by the following formula, which is sufficiently exact for the purpose. In it M intensity of the horizontal component of the earth's magnetism, which in absolute measure (millimètre, milligramme) at St. Petersburg 1.64, E the distance of the deflecting magnetic from the pivot of the compass 300 mm., and V = angular deviation produced.

=

=

M = H.E3 tan V = 221 105 tan V.

=

From this the reduced value per unit of weight (in this case 1 gramme) is obtained by dividing the value found for M by the weight of the bar in grammes. This affords a standard of com