illuminated with a brightness which only slightly diminishes towards the more refrangible side. A very feebly illuminated field appears then at 66° 54'; it is of small breadth. 12. If the condenser be placed upon Holtz's machine, the continuous spectrum changes at one stroke into the line-spectrum: groups of lines stand out in places which were previously dark; the bright fields split up; and on the field (2), for instance, bright lines start out right and left of the brightest part, while the brightest part itself becomes dark. The colour of the light becomes bluish green. The positions of the individual groups of lines are, from several concordant measurements, the following: 1st group of lines, left boundary right boundary 63 11 20 63 19 30 The right boundary is formed by a very bright line about 3' distant from the preceding. 2nd group of lines, left boundary 63 47 30 In the middle a very large bright double line. 63 58 00 64 8 40 This group is formed from the green field which was before designated as (2); it gives the impression that the brightest part has been torn asunder at the left limit and separated into individual lines. 3rd group, of six lines between 64° 37' and 64° 46' The right boundary is the brightest. 65 4 40 from 65° 40' 10" to 65 44 00 5th group of lines, in the blue, three bright 6th. The bright field which, without the condenser, begins at 66° 18' 40", disappears entirely when it is added. Phil. Mag. S. 4. Vol. 37. No. 251. June 1869. Instead 2 F of it several lines stand out on each side of this place, which, however, cannot be arranged in groups, and are not of great brightness. 7th. In the violet part of the spectrum there are A bright violet line A feebly bright group 6' broad, from 67° 3' to 67 9 00 A bright violet line at 67 36 30 If the discharges of a small Leyden jar be passed through a Geissler's tube filled with highly rarefied oxygen, just the same spectrum is obtained; with a stronger charge it becomes more brilliant without otherwise changing. When the spectrum obtained with a Holtz's machine is compared with that described in § 9 as obtained with the Ruhmkorff's coil, it is at once seen that both are identical, although, owing to the greater brightness in individual groups with the Holtz's machine, a few lines become visible which could not be seen in the former case. It therefore follows that in this case, as also with hydrogen, three distinct spectra may be obtained with induction-currents, according as the gas in the tube has greater or less density. That this difference in the spectra is solely due to the different temperatures of the gas follows from the experiments with the Holtz's machine. The same considerations which in § 7 led to the continuous spectrum being regarded as that corresponding to the lower temperature, and that consisting of groups of lines as corresponding to the highest temperature, lead here to the same conclusion. The continuous spectrum belongs to the lowest temperature (although it is not seen with gas of great density), because it is formed by the continuous discharge of the Holtz's machine. The spectrum described by Plücker, which with gas of suitable density may also be produced in its essential features with the small Ruhmkorff's apparatus, belongs to a higher temperature. The last mentioned, which is attained with gas of the least density by the aid of the Ruhmkorff's coil and of a Leyden jar, belongs therefore to the highest temperature. III. Nitrogen. 13. In investigating the spectra of nitrogen, Geissler's tubes were filled with dry air, after what Plücker states had been confirmed, that dry air furnishes the same spectrum as pure nitrogen. With air in Geissler's tubes no traces of oxygen-lines are seen; and there is here no difficulty in getting the spectrum free from hydrogen-lines; the tube need only be filled a few times with air which has been dried by sulphuric and phosphoric acids.. Using the same induction-apparatus as in the previous experiments, the current just began to pass through the tube filled with air when the pressure was 94 millims; yet the light was not continuous. A continuous passage of the current only occurred when the pressure was diminished to 64 millims, though the luminous intensity was so small that a prismatic investigation of the light was not possible. On a further diminution of the pressure, the brightness of the light gradually increases; and under a pressure of 46 millims. the luminous intensity is adequate for spectrum-investigation. The less refrangible parts of the spectrum in the red and yellow are barely visible; only from the green is the spectrum distinctly present; most beautiful are the violet parts, which are so characteristic of the nitrogenspectrum. The red and yellow parts occur first under a pressure of 30 millims.; but they are so faint, that the shaded bands which Plücker has described in the nitrogen-spectrum of the first order are at most scarcely perceptible. The green part with its rich shading stands out more; but the blue and the violet are the most beautiful; in them the individual flutings are completely developed. Under a further diminution of pressure by 5 millims., red and yellow come out more, and the beautiful shaded bands of the complete nitrogen-spectrum are visible. Under a pressure of 18 millims. the spectrum is completely developed; it quite corresponds to the description which Plücker has given of it*, and to what a spectrum-tube filled with pure nitrogen exhibits. The brightness and beauty of the spectrum increases as the pressure diminishes; under a pressure of about 5 millims. it is developed most brilliantly, and remains so until the pressure of the gas is less than 1 millim. Only when the pressure is so far diminished that it can scarcely be measured by Sprengel's pump does the brightness become less, the darker parts being first extinguished, and finally only the brightest parts visible. In its appearance the spectrum approximates to one of the second order, without, however, changing into one, for no new bright lines appear. With a simple Ruhmkorff's apparatus, then, only one spectrum can be exhibited in a tube filled with nitrogen; a difference in density is only of influence so far, that the spectrum is more or less complete and appears of greater or less brightness. Using, too, a Holtz's machine without superposed condenser, the nitrogen-spectrum of the first order was seen as with an induction-apparatus. Using the condenser or a small Leyden jar, the spectrum of the second order described by Plücker occurred. * Plücker and Hittorf, Philosophical Transactions for 1865. Even when the exhaustion had reached its utmost limit the appearance was quite unchanged. 14. Nitrogen thus only furnishes the two known spectra; and without using a Leyden jar the first spectrum cannot be changed into the second. Hence there is a considerable difference between the behaviour of hydrogen, oxygen, and nitrogen. With the first two gases the same mode of discharge can yield entirely different spectra in the enclosed gas, according to its density. Hence this difference can have no other reason than the higher or lower temperature to which the gas has been heated, and which, as mentioned in § 7, depends on the different density of the gas. It must be assumed that the emissive power of both gases does indeed essentially vary with the temperature. The case is different with nitrogen: the difference in temperature produced by the different density of the gas is not sufficient to change the spectrum; the mode of discharge must be changed. Nitrogen can only be brought into the condition in which it yields a spectrum of the second order, by the sudden passage of large quantities of electricity, obtained by simultaneously interposing a Leyden jar in the circuit of the induction-coil, or by passing the discharge of a Leyden jar with the Holtz's machine. Hence we may speak of an allotropic condition of nitrogen, which furnishes the second spectrum, and which is formed by the sudden discharge of large quantities of electricity, which, however, returns to the ordinary form as soon as the temperature diminishes. To be sure, no explanation is thus given of the difference in deportment of nitrogen and other gases; this can only be expected from further experiments, which will be reported upon in due course. Bonn, August 1868. LIX. On the Motion of a Palladium Plate during the Formation of Graham's Hydrogenium. By JAMES DEWAR, F.R.S.E.* G RAHAM, in continuing his exhaustive researches on diffusion, has recently examined the relation of gases to various colloid septa. The remarkable discovery of Deville and Troost of the permeability of platinum and iron by hydrogen at a red heat, he has expanded into a general examination of the relative rates of passage, at high temperatures, of the various gases through different metallic septa. Further, he has proved that different metals have a specific occluding power over certain gaseous elements, retaining them in combination at low tempe ratures, although the absorption took place at a red heat. Of * Communicated by the Author, having been read before the Royal Society of Edinburgh, March 1, 1869. the many astonishing discoveries made during the course of these investigations, probably the most remarkable is the occlusion of hydrogen by palladium. This metal, whether in the form of sponge or hammered foil, when heated and cooled in an atmosphere of hydrogen, absorbed between six and seven hundred times its volume, increasing to the enormous occlusion of 982 volumes when the metal used had been deposited by voltaic action. This occlusion of hydrogen, Graham has shown, can be easily effected at low temperatures by making palladium the negative electrode during the electrolysis of water. He has also shown that the metal charged with hydrogen increases greatly in volume, and that its physical properties are entirely modified. So marked is the change in the physical, electrical, and magnetic properties of the combination, that the only class of compounds we can compare it with are the metallic alloys. In the occluded state the chemical intensity of hydrogen is increased, many reactions being effected by its agency beyond the power of the element in the free state. Graham, as a general result of his experiments, considers the occluded gas to exist in the form of a solid, with all the physical properties of a metal. During the course of an experimental exhibition of Graham's discovery, I noted several phenomena associated with the occlusion of hydrogen by palladium when it is made the negative electrode during the electrolysis of water; and as they illustrate in a new form the results already arrived at by the Master of the Mint, with his permission I am induced to bring them before the Society. If a palladium plate, used as the negative electrode during the decomposition of water, be arranged at right angles instead of parallel to a similar platinum plate, the hydrogen in a short time is evolved at the edge of the palladium plate nearest to the platinum electrode, no trace of hydrogen coming from any other part of the plate. Gradually, as the saturation takes place, the hydrogen seems to travel slowly along the plate, and only after saturation is it freely evolved from the whole surface of the electrode. If we now reverse the current, so as to evolve oxygen at the palladium plate, immediately the nearest edge begins to evolve gas, the rest of the plate remaining tranquil; the evolution of oxygen moves along the plate in a gradual manner. This gradual transference depends on the time necessary to effect the occlusion, and on the relative intensity of the lines of force. When a palladium plate charged with hydrogen is brought into contact with a platinum electrode freely evolving oxygen, evolution of gas is immediately arrested over the entire surface of the electrode. The same plate, free from hydrogen, when brought into contact with a platinum electrode evolving hydrogen, |
