platinum (?HgO). August 7th reading 63, absorption 5 mm. Started sparking 10.15. Stopped 12.15, 63 mm no change.' After removing the oxygen with alkaline pyrogallol, the volume of the gas was found to be 8456 of the original volume. Supposing that gas which had disappeared to be nitrogen, Ramsay proceeded to calculate the density of the residue. He expresses the result: Density of Q. 20.01. However, it was impossible to check this result immediately by purifying the bulk of the gas. Disaster temporarily brought the work to a standstill. He writes: "On arriving at the laboratory this morning, Williams was there and showed me the reservoir tube broken clean off at the top, not at the seal. It must have gone of its own accord, as no one had been in the room. All gas of course lost." On the same day he received the following letter from Lord Rayleigh. "DEAR PROF. RAMSAY, "I believe that I too have isolated the gas, though in miserably small quantities. When I spark away (after Cavendish) 50 c.cs. of air with oxygen added as required, I get a residue of 3 c.c. which is neither oxygen nor nitrogen (no hydrogen). "The same operations conducted with 5 c.c. of air give a very small approximately proportional residue only. I have estimated the gas as but 1/150 or less of the nitrogen, and the density correspondingly higher than what you give. I have concentrated X by diffusion, collecting at the end of a tobacco pipe 1/30th of the gas which goes through the pores into a vaccum. The air so prepared seems to contain twice as much X as ordinary air. I am preparing to develop this method further. My attempts to accumulate larger quantities than 1 c.c. have only partially succeeded, I think because of the solubility of X in water, but I am trying sparking on a larger scale in London. However I c.c. was sufficient to allow of the spectrum being observed at atmospheric pressure between platinum points. Like you I could find no new line as I had hoped, but the nitrogen lines were conspicuously absent, or only very faint. "As to publication, I had thought of giving at Oxford some definite results of the work (with urea, etc.) undertaken to settle the question of the unity of chemical nitrogen, and perhaps throwing in such results as I have from the repetition of Cavendish. But it seems now so mixed up with your work as to be difficult or impossible to treat separately. My own feeling is that the only solution is joint publication. Doubtless your last results go further than mine, and are probably better established. But as you suggest the whole is founded upon work which I had carried up to a certain point and was continuing. If this course is to be adopted, the question arises whether anything had better be said as yet. If not, I would keep back my further results as to chemical nitrogen. I shall be at Oxford staying with the Warden of Merton, and I shall be glad to hear your views. "Yours very truly, RAYLEIGH. The method of effecting the combination of oxygen and nitrogen by passing sparks through a mixture of the gases was first described by Priestley in his book entitled Experiments and Observations on Different Kinds of Air. He found that the air so treated diminished in volume and became "noxious." However, it was Cavendish who first made an accurate study of the phenomenon, which is therefore associated with his name. His apparatus consisted of a glass tube bent into a flat A. The tube was nearly filled with mercury, and when inverted and placed with the limbs in two cups filled with mercury, the mercury remained at the same height in either limb, and air filled the apex. The mercury in one cup was connected with the conductor. of a frictional electric machine, and the other was earthed; then on passing sparks from the top of one mercury column to the other through the air in the tube, a change was produced in the air which made it turn blue litmus red. Also, soap lees, that is caustic potash evolution, diminished the volume. When oxygen was added to the air in proportion of five parts of oxygen to three parts of air, almost the whole of the air could be made to disappear. He showed that the product was nitre. Finally, air was sparked with excess of oxygen till no further diminution of volume was observed. A solution of liver of sulphur was then let up into the tube, and this absorbed the oxygen, leaving a bubble estimated at 1/120 of the nitrogen of the original air. The story, with quotations from the author in the quaint language of the phlogiston theory, is told in Ramsay's Gases of the Atmosphere. The apparatus used by Lord Rayleigh at Terling Place in his first repetition of the Cavendish experiment is shown in Fig. 5. By means of it he succeeded in producing only what he called “a miserably small quantity" of the gas. However, the experiment served the purpose for which it was designed, that of demonstrating the correctness of Cavendish's statement. It also provided an invaluable method for removing nitrogen, traces of hydrogen, etc., from small quantities of argon, helium, etc. Modified only to the extent that the gas was confined over mercury instead of over water, it was used by Ramsay and myself throughout the research on the rare gases for the final purification of samples of gas for the determination of physical properties; the actual procedure was as follows. After contraction had ceased, the spark points were removed from the tube, and the gas was transferred to a gas pipette with the usual inverted syphon arrangement, taking care to take none of the alkaline solution into the pipette. A clean dry tube was then taken and filled with mercury, and while standing in a bath of mercury a button of phosphorus was introduced into it, the tube being tapped till the phosphorus rose to the top. The tube was then warmed, when the phosphorus melted and stuck to the inside of the glass. On introducing the gas the phosphorus burned with the oxygen, the phosphorus pentoxide formed absorbing any moisture in the gas. Accidents occasionally happened, but very rarely; yet it was anxious work when it came to handling small quantities of which were absolutely irreplacable. gases In the original experiments the gas was confined over weak soda solution in the tube A, standing in the vessel B. DD are platinum wires sealed through the closed ends of the U-tubes CC, and joined to wires leading to the terminals of an induction coil, by means of which a spark was maintained between the platinum points. Starting with say 50 c.cs. of air, to which oxygen had been added in excess, sparking was 'continued, with addition of more oxygen, till there was no visible contraction after an hour's sparking. The gas Phil. Trans., 1895, Vol. 186, p. 198. was then transferred to "a small measuring vessel FIG. 5. sealed with mercury," and on treatment with alkaline pyrogallol the gas shrank to 0.32 c.c. A further 50 c.cs. of air was added to this residue, the process was repeated, and a final residue of 0.76 c.c. was obtained. Several repetitions of the experiment showed that the quantity of residue was proportional to the quantity of air used. It is altogether unprofitable to discuss the question as to who suggested the use of the oxygen or the magnesium method of removing nitrogen from the air. In 1894 the combination of oxygen and nitrogen in the electric spark, and the combination of nitrogen with magnesium when the metal was burned in air were commonly demonstrated as lecture experiments. It was usual to burn a piece of magnesium ribbon-it was called "captive sunshine" when, as a boy, I first remember buying a piece of it in about the year 1884-the oxide was collected in a porcelain dish, and when moistened with water, smelled strongly of ammonia. Actually the first suggestion to use magnesium for the analysis of air was made by a Greek student, Aslanaglou, of my year at University College. Writing to the Chemical News on August 29th, 1890, he describes the product of combustion of magnesium in air as a mixture of nitride, hydrate, and oxide, continuing: "It is evident, therefore, that the quantitative amounts of nitrogen and aqueous vapour in the hydrate will vary with the different atmospheres formed in the different towns, countries, districts, etc." We did not feel particularly proud of our fellow student's performance, and I do not doubt that our remarks about the letter were more pointed than polite. However, the suggestion passed into oblivion, or doubtless some of Ramsay's contemporaries would have named Aslanaglou as the real discoverer of argon. Attempts were made to establish claims which were equally slender. Cavendish's experiment seems to have been generally well known to chemists of Ramsay's generation, when the field of knowledge was rather less. crowded than it is to-day, and men had leisure to enjoy the study of the work of the pioneers of chemistry. The names of Ramsay, Dewar, and Crookes have all been mentioned as having made the suggestion to Lord Rayleigh that he should use this method. I don't think that Ramsay ever made the claim on his own account, but he certainly knew of Cavendish's work, and there is a tradition in the laboratory where I am now again working in the University of Bristol that he once proposed the investigation to a Bristol student. As I have already stated he actually used the method for the first time in the interval between writing to Lord Rayleigh, on August 4th, and receiving his reply. L II A NEW GAS ORD RAYLEIGH'S experiments on the densities of the principal gases had proved beyond question that atmospheric nitrogen was heavier than chemical nitrogen by a constant amount, whatever the source from which the latter was prepared. His repetition of the Cavendish experiment proved that the gas in atmospheric air, which could not be made to combine with oxygen, was a constant quantity. Ramsay's experiments had led to a similar conclusion. They also showed that the residue obtained after removing the nitrogen from the air had a density very close to 20, that of oxygen being taken as 16. The combined results pointed to the extraordinary inactivity of the new gas; and considering the facts as they appear to-day, no one can possibly question the propriety of making an announcement of such an unquestionable discovery at the meeting of the British Association which assembled at Oxford a few days after the work which I have described had been brought to a conclusion. The record of the Meeting of Section B in the Report of the British Association is very brief: "Monday, August 13th. Lord Rayleigh, Sec. R.S., and Professor W. Ramsay, F.R.S., gave a preliminary account of a New Gaseous Constituent of Air." There is a concise, and probably accurate account of the proceedings in Nature: "On Monday a large audience assembled to hear the announcement by Lord Rayleigh and Professor Ramsay of the existence of a new gas in the atmosphere. It appears that certain experiments of Cavendish pointed to the presence in the air of some substance other than the gases with which we are familiar. Attention was recalled to this substance by the fact that the density of nitrogen obtained from atmospheric air differed by per cent. from the density of nitrogen obtained from other sources. It was found that if air (with excess of oxygen) be subjected to electric sparks, the resulting nitrous fumes. absorbed by potash, and the excess of oxygen by alkaline pyrogallol, there remains a residue which is neither oxygen nor nitrogen as can be seen by the spectrum. The same gas can be isolated by exposing nitrogen from air to the action of magnesium. As the magnesium gradually absorbs the nitrogen the |