mercury-reservoirs D and E, the adjustment of which in a higher or lower position gives the necessary pressure for the transfer of gas or liquid in either direction. In order to ensure uniformity of temperature during the progress of an analysis, the eudiometer is waterjacketed, and as it would be difficult to adjust the vessel E so that the level of the mercury in it was precisely that of the mercury in the eudiometer, a comparison-tube G is provided. The eudiometer is calibrated from the point where the capillary tube opens into it at the top to a mark at the bottom bearing the number 100. This space is 50 cubic centimetres, but it is divided into 100 parts for the convenience of reading off directly percentages by volume. The apparatus is filled with mercury, thus avoiding the errors which are liable to creep in when a gas-apparatus is worked with water or a saline solution. The gas is measured in the eudiometer. The absorbents are drawn into the laboratory-vessel, followed by the gas. After absorption has taken place, the gas is returned to the eudiometer for measurement, and the reagent is expelled at the point P. After a reagent has been used in the laboratory-vessel, water can be introduced into it in order to wash it-and the mercury contained in it quite clean before any mercury is sent from the laboratoryvessel to the eudiometer, thus keeping the latter clean. Moreover, after an analysis has been made, the eudiometer, the laboratoryvessel, and the tube connecting them, can be easily washed out with dilute acid before another gas-sample is introduced.

It is of the utmost importance that the apparatus should be kept as clean as possible. The mercury, when dirty, should be cleaned by shaking it up with finely powdered cane-sugar and filtering it through a pin-hole in a cone of filter-paper, or by some other method. The stop-cocks A and B require lubricating: only castoroil should be used for this purpose, and of that as little as possible. Should the inside of the apparatus become greasy, it can be cleaned by running methylated spirit through it.

Method of Using the Apparatus.—Where a water-supply is available, the inlet M is connected with the supply and the outlet N with the waste. In the absence of a flow of water, M is plugged and the jacket is filled with water at N. The stop-cock B is closed, a small beaker is placed under P', E is filled with mercury, and the tap A is turned so as to open the passage from the eudiometer L to P. As soon as the mercury commences to drop from P, the tap A is turned off. This operation having been repeated for the other side, by pouring in mercury at D, the entire system is filled with mercury. The nozzle P is then connected with the gas to be analyzed; vessel E is lowered and tap A is turned to allow the gas to flow through P

into the eudiometer L until the level of the mercury is below the 100 mark; tap A is then turned off. The tubing having been disconnected from P, the height of vessel E is adjusted until the level of the mercury in the eudiometer is two divisions above the 100 mark. By pinching the rubber tube leading from E to the eudiometer, and momentarily opening A, the excess of gas is allowed to escape. A is turned so as to allow a little mercury to flow from C to L, and is then closed. The quantity of gas which was in the capillary tube between A and the top of the eudiometer having thus been driven over into the measuring-vessel, which previously contained 99.6 volumes, will make up 100 volumes. The cock B must now be opened, and the height of E adjusted until the levels of the two mercury-columns in the eudiometer L and the side tube G are exactly the same. If the operations have been performed properly, the eudiometer will now contain exactly 100 volumes. The precise point above the 100 mark at which the mercury must stand so that after the gas in the capillary tube at the top has been driven over, 100 volumes may result, must be found by trial, as the capacity of the capillary tube may vary slightly in different apparatus.

To make an absorption, a small beaker containing the absorbing liquid is placed under P, D is lowered, and the stop-cock A is turned so as to draw the required quantity of liquid into the laboratoryvessel C. When the desired quantity has been admitted, A is closed. B is then closed, E is raised, and A is turned so as to allow the gas to pass from L to C, followed by a small quantity of mercury. By closing A and raising D the gas in C is compressed, and when the absorption is complete E is lowered and A is opened carefully to permit the gas to return to the eudiometer. When the absorbing liquid reaches the capillary tube leading from the laboratory-vessel C, the flow must be carefully regulated and A must be turned off as soon as the liquid has just reached the point where the capillary tube joins the stop-cock. A is then turned to allow the rest of the liquid to flow out from P into a beaker placed for its reception. By lowering D, a small quantity of clean water is drawn in to wash the mercury in C. This water is driven out again and a small quantity of mercury is sent over from C to L to fill the capillary tubes. By opening B the levels are adjusted as before, and the reading is then taken.

Determination of Carbonic Acid.-A small quantity of a fairly strong solution of caustic potash will effect complete absorption of the carbonic acid in a few minutes.

Determination of Oxygen.-A small quantity of pyrogallic acid

solution is introduced into the laboratory vessel, followed by sufficient potash solution to make it strongly alkaline. The absorption can be hastened somewhat by moving the mercury-vessel D up and down. When the absorption is thought to be complete, and the gas has been transferred to the eudiometer, the laboratory-vessel is washed out with water, and the gas is tested with a small quantity of fresh pyrogallic acid and potash mixture; any trace of oxygen left will at once become evident by its blackening action on the reagent.

Determination of Carbonic Oxide.-This gas can be absorbed by means of a solution of cuprous chloride in ammonia. After the completion of the absorption, the laboratory-vessel should be washed out with some weak ammonia solution, then a small quantity of dilute sulphuric acid solution drawn into it, and the gas returned and left in contact for a few minutes, in order to free it from ammonia-vapour, before measurement.

Fig. 2.

At the end of each analysis the laboratory-vessel and the eudiometer should both be washed out with dilute acid. For this purpose a 5-per-cent. solution of sulphuric acid, coloured with a few drops of methyl-orange solution, may be conveniently used. This solution is pink while acid, but becomes yellow when alkaline. In the course of working, it will be found that the mercury is liable to accumulate unduly in one of the two vessels D and E. When this occurs, the vessel can be lifted from its containing box, and some of the mercury poured into the vessel on the other side. If it is desired to obtain results with the utmost accuracy, and there is reason to suspect change of temperature in the circulating-water, a small trap fitted with a thermometer, as shown in Fig. 2, can be introduced into the exit-pipe N.

Scale 3 Inches - Foot 2.

Foot

The Paper is accompanied by a drawing, from which the Figures in the text have been prepared.

OBITUARY.

ROBERT JOHN BILLINTON, Locomotive, Carriage, and Marine Engineer to the London Brighton and South Coast Railway, died at Brighton on the 7th November, 1904, in his sixtieth year. Born at Wakefield on the 5th April, 1845, he served a pupilage to Messrs. Sir W. Fairbairn and Sons, Manchester, and to Messrs. Simpson and Company of Pimlico, and subsequently acted as Assistant Manager to Messrs. Walker and Eaton of Sheffield. In 1870 he was appointed assistant to the late Mr. W. Stroudley, then locomotive superintendent of the Brighton Railway, with whom he remained until 1874, in charge of the design and construction of locomotives and rolling-stock, when he transferred his services in a similar capacity to Mr. S. W. Johnson, locomotive superintendent of the Midland Railway at Derby. He occupied this post until 1890, when, on the death of his former chief, he succeeded to Mr. Stroudley's appointment on the Brighton Railway, a position which he retained until his death. The powerful bogie express engines and bogies carriages now running on that line were introduced during his tenure of office.

Mr. Billinton was a Member of the Institution of Mechanical Engineers, the Iron and Steel Institute, and the Institution of Naval Architects.

He was elected a Member of this Institution on the 10th January, 1899.

THOMAS PROCTER CAMPBELL, born on the 7th May, 1841, was trained in an architect's office, but subsequently adopting the profession of engineering, he obtained an appointment on the East Indian Railway in 1858.

Arriving in India in December of that year, he was posted to Pandu Nadi, about 50 miles above Allahabad. After 2 or 3 years on the Allahabad-Cawnpore section of the line, Mr. Campbell joined Mr. Brookes, then Resident Engineer of the Tonse Bridge, near Mirzapur, and remained on that work till its completion, when he was transferred to the construction of the line from Allahabad to Jabalpur, and placed in charge of the

Hiran bridge, three 110-foot spans of lattice girders on well foundations, about 21 miles from Jabalpur.

On the completion of the line into Jabalpur, Mr. Campbell was employed on the terminal station works, where he remained till the summer of 1868, when he came home on furlough. On his return to India in 1869 he joined the Oudh and Rohilkhand Railway, then under construction. He remained 7 years on this line, and was promoted to Resident Engineer, his last work being the construction of the central workshops at Lucknow. In 1876 Mr. Campbell returned to the East Indian Railway, where he spent the remaining 20 years of his service, becoming successively Resident and District Engineer. At the date of his retirement in 1896, he held charge of the Delhi District.

Mr. Campbell's record, whilst it does not include any works of outstanding importance, is one of quiet persistent work, and duty performed with faithful and unsparing zeal. To his rigid adherence to duty under severe climatic conditions up to the end of his service must be attributed the constant ill-health from which Mr. Campbell suffered after his retirement, and which culminated in his death at Dalhousie, Punjab, on the 3rd May, 1905, in his sixtyfourth year.

He was elected an Associate Member of the Institution on the 1st December, 1868, and was transferred to the class of Members on the 30th April, 1889.

HENRY WHEELER DAVIS, who died on the 3rd April, 1905, aged 76, was the youngest son of Nathaniel Davis, for many years Treasurer of the old Eastern Counties Railway. The subject of this notice commenced his career under the late Mr. James Samuel, then Resident Engineer of that line, and subsequently acted as assistant and principal assistant to his successors, until in 1866, on the retirement of Mr. Robert Sinclair, Mr. Davis was appointed Chief Engineer of the system, which before that date had become by amalgamation the Great Eastern Railway. He held this appointment until his retirement in 1873, when he engaged in private practice, which, some years before his death, he was obliged to relinquish on account of failing health.

Mr. Davis was elected a Member of the Institution on the 15th May, 1866.