within one or two in a million. After standing for nearly three weeks they were marked all over, and varied nearly a tenth per cent. among themselves, and more than that compared with 1,621, which had been in another solution acting as a standard in other experiments.

From this it appears that in order to give good results the zincs should be heavily amalgamated. If thoroughly amalgamated, practically any zinc will do. The zincs which gave low E. M.F.'s were always the small ones-that is to say, those which would get the thickest coatings of iron.

To find out whether amalgamating altered the readings, an amalgam was made by touching mercury with zinc. It was more than per cent. wrong. After dissolving more zinc, the reading was still low, though by stirring in zinc it soon came up to 0.9984 and 0.9994. A series of amalgams was made containing 25, 12, 5, 2, and 1 per cent. of zinc and one with less zinc. These were put in solution 1,628 of zinc sulphate with little iron. The solution was poured in and out many times to secure homogeneity, an important matter, to which I will return. The readings obtained

The second readings, which were taken a few minutes later, are close enough. Amalgams therefore having some 24 per cent. of zinc will do. Lord Rayleigh has done a great deal of work on amalgam cells, and he found the amalgam always cracked the tube. I made one H cell with amalgam in 1888 which had the platinum led out at the bottom, but it did not crack. I made several other cells at the same time with an internal test tube to hold the amalgam, and none of them cracked. On the other hand, the 5 per cent. amalgam broke its test tube. I would suggest that platinum sealed to German glass is always left under constraint, and the crystallisation of the zinc also adds pressure which cracks the tube. My H cell had a little lead glass just round the platinum; moreover, it was very small. The 24 per cent. amalgam is liquid, so there is no danger from it.

To sum up as to the zinc, then, I would suggest that if rod be used it should be well amalgamated first. It is gradually amalgamated in the cell, but this process is slow, especially if the solution contains basic zinc sulphate. The gradual amalgamation is probably one process that makes many cells take a few weeks to arrive at final values. If zinc rod is used, it would seem better to use platinum, silver, or zinc wire to make connection, and to avoid solder, otherwise it is only a matter of time till the zinc falls off. All these troubles are swept away by the use of zinc amalgam, as proposed by Lord Rayleigh. There is another reason in its favour to be mentioned under the head of temperature coefficients.

THE ZINC SULPHATE SOLUTION.

The chief difficulty arises from the presence of iron. Samples were bought from Messrs. Hopkin and Williams and Becker. These both contained iron. Iron is often to be seen as a deposit after shaking up with zinc carbonate. Some sulphate of zinc was made by dissolving "pure" zinc in redistilled sulphuric acid. Provided there is always excess of zinc, one would expect little iron. Another batch was made by leaving excess of commercial granulated zinc in dilute commercial acid, which itself contains iron. Neither of these had enough iron to blacken zinc. None of the samples of zinc sulphate blackened copper. When I say that samples of sulphate of zinc bought as pure from such people as Messrs. Hopkin and Williams contain iron, I mean that they contain enough to cause variations in cells, not enough to get precipitates of Prussian blue. Zinc or barium carbonate might be expected to precipitate all the iron. It does not, however. Probably carbonate of iron really exists and remains in solution. In experiments on alkaline secondary cells with iron plates, I produced a green compound which appeared to be a real carbonate of iron. The best method of dealing with iron is to acknowledge its presence and amalgamate the zincs. The best zinc sulphate solutions I used were obtained by dissolving granulated zinc in dilute commercial acid, keeping the zinc in excess. Solution 1,628, already mentioned, was made this way. Of course, pure zinc and acid may be used. This method produces basic sulphate, too, especially if hot. This can be afterwards neutralised by acid.

It is usual to shake up zinc sulphate with "carbonate," and to filter before use. This is probably because the sulphate reddens litmus paper. Neutral zinc sulphate reddens litmus paper, however, so there seems little reason for adding zinc "carbonate." I imagine that any salt which can give up a little acid by forming a basic salt, will redden litmus. The sulphate can be neutralised by being left in contact with zinc. A solution that is strongly alkaline to methyl orange reddens litmus paper. What is bought as zinc "carbonate," is either a basic carbonate or a mixture containing oxide. There are some half-dozen basic sulphates of zinc, and shaking up with "carbonate" forms some of these. presence of basic sulphate in the solution prevents the mercury salt getting to the zinc to amalgamate it. Mercurous sulphate is, I believe, split up by water into basic and acid mercurous sulphate. Basic zinc sulphate precipitates any normal mercurous sulphate that may be dissolved, and the acid sulphate as basic mercurous sulphate or as mercurous oxide. This accounts for the dark grey deposit often seen on the top of the paste in cells prepared in the usual way. The mercury salt thus cannot get to the zinc till the

whole of the basic zinc salt has been used up.

The

In order to test the effects of various solutions of zinc sulphate, test tubes with feet were filled with the solutions, and were connected with the standard solution by a syphon tube. This had one end blocked with filter paper to prevent mixture of solutions. It is needless to give tables of readings. Anomalous results were obtained at first, but they were all traceable to the zincs. After amalgamation, all the saturated solutions gave the same readings within a few in a million. 1,621, made by dissolving pure zinc in acid, was 100 in a million too low.

To test the effect of acid, I added a few drops of "normal"-that is, very dilute sulphuric acid-to a saturated solution of Hopkin and Williams's zinc sulphate. The solution was then strongly acid to methyl orange. The solution read 1001200 against zinc, 1,621 in some more of the same solution. On neutralising with zinc carbonate, the reading was 1001100. Free acid appears to make little difference. This experiment was repeated with new solutions. The two zincs were first tested in the same solution. One gave about eight in a million; this was moved into the other tube, B, of the same solution and taken as a standard. The reading was now 1000110 for A. Five drops of pure acid were put in a small test tube, and a little of A added, and the tube cooled. Then the acid was added to A, which became strongly acid to methyl orange. The reading was now 1000000. The liquid was then shaken up with zinc carbonate till alkaline to orange, and the reading rose to 1.0009. On the other hand, a solution which had been left in contact with excess of zinc carbonate for some weeks gave no deflection. In any case, the zinc will eventually use up any free acid. These anomalous results may be partly due to slight differences of density in the solutions, and partly to unascertained causes.

This is the last point that need be mentioned in connection with the solution. The solubility of zinc sulphate varies enormously with the temperature. If a tube of it is left with crystals at the bottom, when it comes to a low temperature all the crystals collect at the foot. When the temperature rises, the upper part can only become saturated again slowly by diffusion. To test this, two zincs were put in a test tube of saturated solution with crystals at the bottom. After a few days the top zinc gave 10005 compared with the bottom. The lower zinc was then raised, reading 1. The other zinc was then depressed, reading 0.999475. The difference of density thus causes an error of 5 in 10,000. This tube was some 10 centimetres long, so the error in standard cells may not be so great as this, still, diffusion is very slow even in a short cell. This also influences the temperature coefficient. The question of supersaturation has already been dealt with so thoroughly by Lord Rayleigh that I saw no use in experimenting on it. There are some half-dozen hydrates and an anhydrous sulphate of zinc. If the solution is not heated above blood heat, the Zn SO, 7H,0 crystallises out.

To sum up the chief errors caused by the solution are due to traces of iron and variations of density. The zinc should therefore be amalgamated. I would suggest that zinc carbonate be omitted. To avoid changes of density, if a saturated solution is imperative, a zinc amalgam must be employed with crystals lying on its surface. This can be done in one of Lord Rayleigh's H cells, or a smaller tube can be slipped into the cell. The surfaces must be on the same level, else all the crystals will collect on the lower one, as that lowers the centre of gravity of the system, thus doing work. A small test tube full of crystals was put inside a larger tube of saturated solution, and this action was found actually to take place.

There is some further complication between the zinc and its solution. Even amalgams will give very large variations if moved about in the solution. This is the case even with dilute solutions. I can scarcely hazard a suggestion. Some Lippmann effect may be produced by shaking the amalgam. This does not seem at all likely. It is possible that zinc, even when amalgamated, will dissolve in sulphate, forming basic sulphate, provided the zinc is moved so that the hydrogen can get away in solution, as it cannot come off in bubbles. Ordinary zinc will dissolve in boiling zinc sulphate solution, the solution turning milky on cooling; but I cannot make out whether amalgamated zinc does or not. In the cold it gives off bubbles slowly. I do not know whether any surface boundary layers can cause such effects; whether there is an action like a frictional electric machine going on. MERCUROUS SULPHATE,

Samples of mercurous sulphate were bought from Messrs. Becker and Hopkin and Williams. I also had some samples from Hopkin and Williams, Becker and Griffin, bought in 1888, and some I made myself at the same time. I also prepared some new specimens by heating in an oven a flask with mercury twice distilled in vacuo and redistilled acid. The temperature was kept at 125deg, C. by a mercury regulator, and only a portion of the mercury was dissolved. The sulphate was washed with distilled water and dried. It is in white crystals.

Other samples were made by dissolving mercurous nitrate in water slightly acidified with nitric acid, and precipitating with zinc sulphate. This is perfectly white unless much washed. A yellow colour then appears.

There are several sulphates of mercury.

There is mercuric

a

sulphate, which, on shaking up with water, splits into acid mercuric sulphate, which is soluble, and basic mercuric sulphate, which is a bright yellow insoluble powder. There is, I believe, mercuric-mercurous sulphalte. I do not know it, nor do I know if there are acid and basic salts to correspond. There is also mercurous sulphate, which some say is soluble in water, but which, I believe, splits up into acid mercurous sulphate, which is slightly soluble, and basic mercurous sulphate, which is a dirty yellow powder. In testing the pastes, some of that used for the

B.O.T. cells was used as a standard. Zinc No. 1,621 was kept in a solution of sulphate, and was connected by a syphon with the tube of standard paste, the cell thus made being checked against the B.O.T. standard. After a few days there was never any deflection. The pastes could not be measured so easily as zincs or solutions on account of their high resistance. There is also frequent trouble from want of contact with the platinum wires. Merely heating these and dipping in mercury is not always enough. I found it better to bend the platinum into an eye and to boil it in mercury, and not to consider it amalgamated till it took up a film of mercury with it, or to use sodium amalgam. The syphon tube was no longer stopped with filter paper, so its resistance was lower. Some of the experiments on pastes are given in the following table:

1,650 was B.O.T. paste. Its readings are given in terms of 1,615, the B.O.T. standard. The other pastes are given in terms of 1,650. Columns one, two, and three are not corrected for the variation of 1,650, as from the readings of 1,653, 1,654, and 1,654, it looks as if the B.O.T. cell had been varying owing to alterations of density of the zinc sulphate solution.

1,651.-Paste given to me by Mr. Fox Bourne. The bottle had Becker's label, but the paste was grey like that of Hopkin and Williams, from whom Mr. Fox Bourne had also bought sulphate. It was always far too high.

1,652.-Paste of basic mercuric sulphate. This was always very low. This goes to show that basic mercuric sulphate is harmless, so that mercuric sulphate may perhaps be rendered harmless by washing it thoroughly.

1,653.-Paste made by dissolving mercury as already described. It began low, gradually coming to be too high on shaking. 1,654.-Paste made by precipitating mercurous nitrate with zinc sulphate. This is quite white unless copiously washed, then it is slightly yellow owing to the presence of basic protosulphate. Tap water containing chalk must not be used for washing. This paste contains no mercuric salts. To test the pastes for mercuric salts, they were shaken with water and the solution precipitated with salt and filtered. The filtrate was tested with hydrogen sulphide and with potassium iodide. This paste seem right. 1,655.-Paste bought in 1891 from Becker. This was a white powder, and remained white as paste. Contained no persalt. Gradually fell to normal. The pastes nearly always give high readings after stirring up well.

1,656.-Paste from Hopkin and Williams, 1891. Powder, grey, turning yellowish in the paste, contains a good deal of persalt. This is the paste generally used for cells. Two other specimens bought from the same firm some years ago also contained persulphate. The paste begins very high, gradually falling. Washing out the persulphate may help this. I do not know if sodium formate could be used to reduce the persalt.

1,657.-Paste from Griffin. Grey powder. Contains a great deal of persalt. On July 24th, 1891, excess of zinc carbonate was added to this paste. It frothed up and remained as a sort of sponge. The reading then fell gradually till it was too low. 1,658.-Paste made by slow electrolysis. M. Potier has proposed a cell made by using a mercury anode and cathode in zinc sulphate; but I have not the reference. The zinc amalgam is, of course, too weak unless a great deal of electricity is passed through. The paste formed is the yellow basic mercurous sulphate. The reading was too high first, and fell to nearly normal. Finally a little acid was added on August 2nd, 1891, and the reading was very high. I made other unsuccessful experiments on Potier cells.

1,612.-H cell with 24 per cent. amalgam. It was always low, probably because its solution was saturated at both poles. Of course, as this was a complete cell, it was compared direct with the B.O.T. cell, and as the temperature coefficients are different, and as there are other causes of variation, it is impossible to say which was at fault. Readings like those of 1,654 may be got with pastes, but not with different kinds of cell.

1,674.-Becker paste, 1888, with zinc carbonate.

1,675. Same without carbonate. Probably there was not enough carbonate to convert the whole of the sulphate into basic salt. Lord Rayleigh recommends the addition of a very small quantity of zinc carbonate to the paste. If the paste has persulphate the carbonate will first convert it into basic persulphate, which appears to be harmless. An excess of zinc carbonate will, of course, waste the mercurous sulphate.

1,679.-Complete cell with Becker paste; put in dry. 1,681.-H cell. Becker paste.

1,685.-Paste made by electrolysis in acid solution.

1,688.-Tube cell, with precipitated mercurous sulphate, made in 1888.

It will be seen that the cells seem to tend towards a common value, but that is almost all that can be said.

The Board of Trade circular directs the paste to be rubbed up with mercury. This would tend to reduce the persulphate to protosulphate. Mercury cannot be rubbed up with protosulphate, as it does not oxidise the surface, so the mercury is not "killed," and collects in a globule. The mere fact of the mercury mixing shows that the sulphate is bad.

MERCURY.

To test the effect of impurities, some amalgam containing silver, copper, and zinc was taken. Sodium amalgam was added, and the whole shaken up with salts of lead, iron, antimony, platinum, and tin. The amalgam was then shaken up well with persalts of mercury. The mecuries to be tested were put in little crucibles in a solution of mercurous nitrate. After a few minutes they all agreed.

Of these cells, 1,680, 1,683, 1,689, had dilute solutions, made up of equal weights of water and Zn SO, 7H,O. 1,680 and 1,683 differed only in the mercury paste; 1,683 gave different coefficients explained. If a cell with the zinc at the top has its solution on the two days. The variations in the other cells may be easily round the zinc is no longer saturated. The temperature coefficient saturated at, say, 15deg. C., and is then heated, the solution of the cell depends on its past history in fact, and is lower in spring than in autumn. An H cell, or a tube cell, with crystals of zinc sulphate in contact with the zinc always has its solution saturated at both platas. 1,682 is an ordinary cell, but the zinc is at the top, hence its low temperature coefficient when heated. If left at 30deg. C. for a few weeks and then cooled, no doubt its temperature coefficient would rise to 0.0008 or 80. 1,684 is cell, with zinc at top. 1,688 is a tube cell, with crystals in an H cell with crystals in both legs 1,685 is an electrolysis both tubes, hence its high coefficient. It is thus clear that the temperature coefficient of a cell may vary from day to day according to its treatment, and may, perhaps, vary 100 per cent. between winter and summer. Even a variation between 0008 and 0006 means an error that may in extreme cases amount to something like a half per cent.-that is, it may be outside the limits I have here taken as worth considering. My water bath differed some varied more during the day and night. A new cell would thus be two or three degrees from one evening to another, and, no doubt, saturated at, say, 20deg., while the standard was saturated at, say, 15deg. If the coefficient of one was 0.0008, and of the other 0 0004, the error might easily be 0.0012 due to that alone. To avoid this difficulty the solution must be kept saturated by using the H or tube cell. Some authorities push the zinc into the paste. This may give the same result, but the zinc then amalgamates too much. Sometimes it has beads of mercury hanging to it, and if these are the only surfaces exposed the E.M.F. may be too low. Over-amalgamation also often detaches the zinc by weakening the solder. Moreover, the formation of so much zinc sulphate may use up all the water, thus rendering the cell dry, or producing a less hydrated salt.

It would seem better to use a weaker solution. Prof. Carhart and others have proposed saturating at Odeg. C. It seems much simpler to use a standard solution of sulphate made by weighing. There is no particular advantage in saturating at Odeg. Č. Weak solution cells have the very great advantage of low internal resistance. No doubt, crystals of zinc sulphate also lessen the available surfaces of the metal and lessen the current that can be taken without polarisation. The construction of the H cell is shown in Fig. 1. One leg contains amalgam with crystals on it, and the other paste. Platinum wires are sealed through glass stoppers, and end in little bracelet snaps. The stoppers are smeared with vaseline to prevent creeping. Fig. 2

is a tube cell. A test tube with a foot has a cork with two holes in it. Through one of them the tube with internal wire is led to the bottom. The mercury and paste are poured in through a funnel. The zinc tube is then pushed through the larger hole in the cork; it contains the amalgam, and has three holes in the side. It is strong to resist fracture. Solution is poured in gently above the level of the first hole. Rangoon oil is poured in over the level of the second hole. The third lets the air out. The zinc connection is then put in, with its cork fitting the smaller tube. If the solution is saturated, the inner tube must be long, so that the zinc surface is on a level with the crystals on the paste. If a dilute solution is used, the inner tube is shorter. This form of cell is convenient for immersing in a bath, and the zinc tubes can be taken out and changed, or tried in other cells.

Prof. Carhart mentions the effect of pressure on Daniell cells. I tried exhausting a Clark cell a third of an atmosphere, and raising its pressure, but though that must have affected the tendency of the liquid to evaporate, it made no perceptible difference in the reading.

CITY NOTES.

Brazilian Submarine Telegraph Company.—The receipts for last week amounted to £4,186.

City and South London Railway.—The receipts for the week ending 22nd inst. were £728, against £653 for the week ending 15th inst.

Electric Installation Company.-Letters of allotment in the Electric Installation and Maintenance Company, Limited, for the supply of electricity to the Crystal Palace, Sydenham, and district, have been posted.

The London Platino-Brazilian Telegraph Company announce that the half-yearly coupon on the 6 per cent. debentures, due on September 1st, will be paid by Messrs. Glyn, Mills, and Co., Lombard-street, E.C.

Western and Brazilian Telegraph Company.-The receipts for the week ending August 21, after deducting 17 per cent. of the gross receipts payable to the London-Platino Brazilian Telegraph Company, were £2,812.

New Company.—The Electric Welding Company, Limited, has been brought out at last. The capital has been put at £460,000 in 25,000 ordinary £10 shares and 1,000 founders' shares of £10. Onethird of the ordinary shares are taken by the vendors, and the remainder are offered to the public-that is, the vendors are to have £83,330 worth of shares. The Board is made up of the following names: Sir George Barclay Bruce, past president of the Institution of Civil Engineers; E. Ashmead-Bartlett, Esq., M.P., 6, Grosvenor-street, London, W.; Captain A. H. Chapman, J.l'., Messrs. Clarke, Chapman, and Co., Victoria Works, Gateshead-onTyne; Wm. F. Gooch, Esq., J.P., managing director of the Vulcan Iron Works Company, Limited; Joseph Kincaid, Esq., M.I.C.E., director of the York Engineering Company, Limited. The consulting engineer is Sir Frederick Bramwell, Bart., F. R.S., past president of the Institution of Civil Engineers, 5, Great Georgestreet, Westminster. The prospectus states that "This Company has been formed to acquire the British patents in respect of Prof. Elihu Thomson's inventions for the electric welding and working of metals." One of the welding machines can be seen at work at the Company's depot, Fanshawe-street, Hoxton.

14063. Improvements in telephonic transmitters. Alexander Marr, 70, Market-street, Manchester.

14067. Improvements in moulds and machinery for the manufacture of glass switch-slabs for electrical and other purposes. Dan Rylands and Benjamin Stoner, Shepcote House, Stairfoot, near Barnsley.

14084. Improvements in electrically actuated time-checking apparatus. Joseph Appleton and William Thomas Burbey, 77, Chancery-lane, London.

14089. Improvements in electrical measuring apparatus. Anthony Reckenzaun, 11, Furnival-street, London. 14096. Combination electric lock James Graham Thompson, 23, Southampton-buildings, London.

14097. Improvements in telephone transmitters, and the methods adopted for the transmission of articulate speech, musical, and other sound waves. Herbert Luzerne Todd and John Todd, 13, King-street, Cheapside, London.

14134. Improvements in thermo-dynamic machines, and in apparatus and appliances connected therewith. William Henry Watkinson, 15, St. James's-row, Sheffield.

14150. Improvements in electric fire detectors or alarms. Harwood Morgan, 55, Chancery-lane, London. (Charles Dion, France.)

14171. An improvement in car conductors from overhead for electric cars George Sylvester Grimstone, 28, Southampton-buildings, London.

AUGUST 22.

14217. A new or improved printing telegraphic apparatus. Sacco Gospare and Luigi Giacomini, 54, Castle-street, London.

14234. Improvements in and relating to electric railways Michaelangelo Cattori, 45, Southampton-buildings, London. (Complete specification.)

14236. An improved method of arranging apparatus for generating and distributing electricity by alternating currents. Henry Robert Low and George Edward Barker Pritchett, 41, Beaconsfield-road, Twickenham.

SPECIFICATIONS PUBLISHED.

1882.

4303. Electrical storage batteries. Frankland. (Second edition. 6d.

1888.

9745*. Electric motors. Manville. (Amended.) 8d.

1890.

1306*. Galvanic dry elements. Smith (Hellesen). (Amended.)

6d.

12209. Electric meters. Heurtey and others. 11d.

12330. Thermo-electric batteries. Gouraud (Dickerson). 6d. 13199. Electric telephone systems. Graham. lld.

14693. Galvanic battery. Bayly (Siliceo). 6d.

15328. Heating by means of electric currents. Zipernowski. 6d. 15536 Telephone receivers. Collier. 8d.

15792. Incandescent electric lamps. Mills (Edison). 8d. 16792. Electric arc lamps. De Puydt. 8d.

(Anton

Joseph

Daniel H. Smith, 141,

1891. 1192. Propelling boats, etc., by electrical energy. 7160. Incandescent lamps. Walter. 6d. 11118 Electrical signals for steam vessels. another). Sd.

Büsser. 8d.

Clark (Tucker and

Bath-street, Glasgow. (Complete specification.) 13926. Improvements in electric locomotives and in electric conductors for use therewith. Llewelyn Birchall Atkinson, Claude William Atkinson, and Frederick Hurd, 1, Queen Victoria-street, London.

13927. A unison apparatus for printing telegraphs. Henry Van Hoevenbergh, 1, Queen Victoria-street, London. 13936. Improvements in and relating to electric motors, chiefly designed for use in connection with ventilating fans. Charles Jacob Kintner, 45, Southampton-buildings, London. (Complete specification.)

13942. Improvements in or connected with apparatus for Thomas completing and interrupting electric circuits. Parker, John Harold Woodward, and Edmund Scott Gustave Rees, 47, Lincoln's-inn-fields, London.

NOTES.

Russian Exhibition.—An international exhibition is announced, under the auspices of the Russian Government, at Odessa in 1894.

Lignite. The use of lignite as a substitute for coal, from which great things were expected in Italy, has proved not to have the value claimed.

Melun.-Negotiations are in progress between the Municipality of Melun, France, and the local company for lighting the public lamps by electricity.

People's Palace.-It has been decided to light up the People's Palace at Whitechapel with electric light. Messrs. Slingo and Brooker are to superintend the installation. Dissolution of Partnership. We see it announced that Messrs. Johnson, Matthey, and Co., Hatton garden, the well-known assayists and metallurgical chemists, have dissolved partnership.

Change of Offices.-The Brush Company are, we understand, removing their central offices from Belvedere

road to their premises opposite the Mansion House Station,

in Queen Victoria-street.

Carbon Works.-As will be seen from our advertisement columns, Messrs. Wheatley Kirk and Co. have for disposal a small compact works, with engine, presses, grinding plant, and furnaces for manufacture of carbon, with stock of material, as a going concern.

Guernsey Electric Railway. As the Channel Islands are outside the provisions of the Board of Trade, the supervision of the erection of the line is being carried out by the Special States Committee, who have appointed Mr. Bernard Drake as their electrical adviser.

Testing Cell.-The Mining and General Electric Lamp Company are now making a very small lithanode cell, suitable for testing purposes. The total weight is 2 oz., including terminals and electrolyte. Its capacity when discharged at the rate of 1 ampere is one ampere

hour.

Helmholtz Medal. In commemoration of Prof. Helmholtz's seventieth birthday a medal like that of Copley is hereafter to be struck by the Berlin Academy. It will bear the portrait of the savant, and is to be given every two or three years as a reward of merit to an eminent German man of science.

Sunderland. An application was made by Mr. R. J. Charlton to the Sunderland Town Council for permission to lay a pipe across the street, underneath the carriageway, for the purpose of conveying an electric wire from Mr. Thornton's Theatre Royal to Mr. Laing's bar. The matter was referred to the Electric Lighting Committe.

Electric Dinner Effects. At the dinner given during the recent visit of the French Navy to Portsmouth, some striking effects were obtained at the dinner table in the Jan Van Beers' style, by lighting up the table (which was of glass) underneath, so as to glow in the "red, white, and blue" of the familiar tricolour. The idea was duly applauded.

Cables to Brazil.-France and Brazil are now in telegraphic communication with each other via the Antilles. On Wednesday the first official message was transmitted over the cable congratulating President Carnot on the fact. On the same day the first direct cable between the United States and Brazil, stretching from Punta Rassa, Florida, to Para, was opened.

Paris Telephone.-M. Jean P. A. Martin, member of the Association of French Stenographers, writes to the Times bearing his testimony to the efficiency of the ParisLondon telephone, which he uses every day, but wishing to know why, when telephoning is perfectly distinct between London and Lyons, red tapeism or other cause prevents the use of the line direct to Lyons.

Shoreditch.-On Tuesday the Shoreditch Vestry, Mr. N. Moss, L.C.C., presiding, decided to obtain power to supply electricity within the parish of Shoreditch for public and private purposes; and to memorialise the Board of Trade for a provisional order under the Electric Lighting Acts, 1882 and 1888, to enable the Vestry to supply the parish or any part thereof with electricity.

adopted a resolution of the General Health Committee accepting the tender of the Manchester Edison-Swan Electric Lighting Company for the electric light installation required at the sanatorium, in accordance with specifications prepared, for the sum of £1,312. Mr. Alderman W. Robinson said that they would save £300 by adopting the system proposed.

Salford.-The Salford Borough Council, on Wednesday,

Missing Letters.-Our readers will doubtless have noticed in the daily papers this week the statement that a large number of letters have never reached their destination, and that over 2,000 have been just discovered. Among these were some addressed to this office. If any of our correspondents have sent any important communication which has not been attended to they would do well to communicate again.

Waterfalls.-The Moniteur Industriel states that an English syndicate is visiting France to investigate the various falls of water with a view to purchase. The price of waterfalls is evidently likely to go up now that light, power, and heat can be so easily manufactured out of their hydraulic power. The French papers remind their readers that a good waterfall has now become to all intents and purposes an "everlasting coalfield."

A Fire Detector.-A neat and simple fire alarm has recently been devised, consisting of a stud placed in a porcelain base and surrounded by, but not touching, a knob of fusible metal of considerable surface projecting into the room. The electric circuit is arranged to energise a relay on the slightest contact of the metal with the stud, and the whole is so cheap that they might be placed in every room of a mill or public building at small cost.

Partnership.-Mr. J. Edward Waller, who has for many years been associated with Mr. Joseph Kincaid as his manager, has entered into partnership with Mr. Edward Manville, consulting electrical engineer. Messrs. Waller and Manville will carry on at 39, Victoria-street, Westminster, S.W., the profession of consulting civil and electrical engineers, a desirable combination in the rapidlyadvancing condition of the applications of electricity.

Obituary. We much regret to have to record the death of one of the members of the Old Students' Association, at the age of 22, Mr. George T. S. Müller, who was well known to most of the members of the above-named association, as he was one of the most regular attendants at