Page images
PDF
EPUB

In taking the observations the rays were started and allowed to run for a definite number of seconds, so that the ionization might reach a steady state, both the electrodes E and I being to earth during this time. At the end of this interval the electrode E was insulated and the key W opened simultaneously, and the rays were allowed to run for a given time. The electrodes E and I were thus allowed to charge up for exactly the same time, and under the influence of the same cone of rays. The reading of the electrometer corresponding to the charge on E was observed, and then the quadrants of the electrometer were discharged and again insulated, and the key W was then closed and the deflexion corresponding to the charge on I observed. This served therefore as a very close check on the constancy of the rays during exactly the same time as the measurements were being made in the cylinder AB.

The first gas tested with this apparatus was of course air, in order to see if the results would be in agreement with the previous experiments. Observations were taken over a range of temperature of nearly 200° C., and the results obtained exactly confirmed the previous results which I had obtained. In this case when the density of the air was kept constant, the amount of ionization also remained constant. The change of temperature of the air had no effect whatever upon the amount of ionization produced in it. A set of results obtained is shown in Table II. as a specimen.

[blocks in formation]

From the results obtained therefore by the two methods there appears to be no doubt whatever that, when the density of a given volume of air is kept constant, the amount of ionization produced in it per second by rays of a given intensity is quite independent of the temperature of the air.

[ocr errors]

Measurements on Carbon Dioxide and Hydrogen.

The next gas experimented upon was carbon dioxide. It was treated in exactly the same manner as air had been. It was prepared in the ordinary way by the action of pure hydrochloric acid on marble, and was dried by passing it over pumice-stone moistened with strong sulphuric acid before it entered the cylinder. The results obtained with this gas were quite in accordance with those obtained for air. Observations were taken over an even greater range of temperatures than in the case of air, and the same law was found to hold with the carbon dioxide as was shown to hold in air. A series of results for this gas is shown in Table III. TABLE III.-Carbon Dioxide.

[merged small][merged small][merged small][merged small][merged small][ocr errors][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small][merged small]

A considerable number of experiments were also made onhydrogen. It was found very difficult to obtain really satisfactory results with hydrogen on account of the great difficulty experienced in making the cylinder perfectly gastight for this gas. The joints which proved so satisfactory in the case of air and carbon dioxide gas did not serve the purpose so well in the case of hydrogen, on account of the greater ease with which it diffuses, and I was consequently never able to make the vessel perfectly gas-tight for hydrogen. The difficulty was also increased by the fact that the amount of ionization produced in hydrogen is so much smaller than in the other gas, and therefore the effect to be measured was much smaller. Consequently the results obtained for hydrogen were not so accurate as those obtained in the case of air and carbon dioxide. However, taking the results as a whole, I think we may conclude that the same law holds for hydrogen as holds in other cases. Table IV. contains a set of results

obtained for hydrogen.

[blocks in formation]

Although there is considerable variation in the above numbers, still I think this want of constancy can be accounted for chiefly by the leakage in the vessel. However, from all the results obtained, I think we may safely conclude that hydrogen follows the same law for the effect of temperature on the ionization as has been established for air and carbon dioxide.

Discussion of Results.

The experiments described in this paper have proved conclusively that in a given volume of gas, kept at a constant density, the amount of ionization produced by Röntgen rays of a given intensity is independent of the temperature of the gas. Prof. Perrin, in the paper to which reference has already been made at the beginning of this paper, described some experiments which he made on this same question. The results which he obtained are, however, not in agreement with those which I have just given in this paper. The method which he employed was a differential one. He made the rays to pass simultaneously between two sets of parallel plates. He then balanced the effect produced on one set of electrodes against that produced on the other set, so that the resultant effect produced on the electrometer was zero. This balance was adjusted at a certain temperature, and then the temperature of the gas in the vessel containing one set of plates was varied and the balance then tested. For the range of temperature from -12° to 148° C. he found no appreciable alteration in the balance. He concludes therefore that, since no change in the balance occurred and the density of the air varied inversely as the absolute temperature, the total ionization would be proportional to the absolute temperature in air maintained at a constant density. I cannot account with

certainty for the discrepancy between his results and those which I obtained, but I think it probable that the apparatus which he used may not have been sensitive enough to detect the alteration in the ionization. He mentions that at the higher temperatures the "heated vessel appeared less active," but attributed this to some other cause. It looks extremely probable that this was a genuine effect, but that the detecting instrument was not sensitive enough to show the effect to a sufficiently great extent. In the experiments which I did the electrometer used was a very sensitive one, giving a deflexion of about two thousand scale-divisions for a difference of potential of one volt between the quadrants. There was therefore no difficulty in detecting any alteration that might take place in the ionization.

In conclusion I desire to express my thanks to Prof. Thomson for the kindly interest shown and advice given throughout the course of this investigation.

Cavendish Laboratory, Cambridge, Nov. 5th, 1903.

XII. Investigation of the Arc in Metallic Vapours in an Exhausted Space. (Contribution from the Research Laboratory of General Electric Co., Schenectady, N.Y.) By E. WEINTRAUB, Ph.D.*

Plates III.-XI.]

Introduction.

HILE the carbon arc has been the object of a great number of investigations, the arc between mercury electrodes, which offers much simpler relations from a theoretical point of view, has so far been investigated but very little. We may distinguish between two distinct periods. In the first one the investigation was limited to the mercury are in air. Thomas Way (1857), (1861)† was probably the first to publish observations on a mercury arc in air. One of his electrodes consisted of a stream of mercury, the second one of a mass of mercury or a piece of carbon, upon which the stream was caused to flow.

Beginning at that time, a number of inventors occupied themselves with the arc, and a great number of patents were taken out on this subject. They were all founded on the same principle as the lamp of Way, and did not contribute anything essentially new to the art or science.

Communicated by the Author.

Dingler's Polytechn. Journ. vol. clvii. p. 399 (1860); vol. clix. p. 46 (1861); also U.S. Pat. 3345, Oct. 8 (1861); Eng. Pat. 2841 (1857).

A certain advance is represented by the patent of Rapieff*, who used a closed vessel with two mercury electrodes, and started the arc by bringing the electrodes into contact and separating them. He used a cooling chamber to condense the vaporized mercury, and speaks of the advisability of exhausting the tube in which the arc is to play.

The study of the mercury are in an evacuated space begins, however, with Arons, who published his observations in Annalen der Physik, vol. lviii. p. 73 (1896), also in Verh. Phys. Ges. Berlin, p. 55 (1892), and Zeitschr. für Beleuchtungswesen, Aug. 15, 1895. From these articles we may date the second period in the development of the mercury arc. Arons described the general properties of the mercury arc, of the electrodes, and even constructed a small mercurylamp which can be used as a source of light, especially for laboratory experiments. Arons also investigated, to a certain extent, the behaviour of amalgams in the tube. The work of Arons will be mentioned many times in the course of this article.

Gumlich published a few observations on the use of cadmium amalgam as an electrode in Wied. Ann. vol. lxi. p. 401, (1897). Further work on the mercury-lamp, especially the method of starting the lamp by an inductive high-voltage shock, has been done by Peter Cooper Hewitt.

This publication contains some results of the work carried out in the Research Laboratory of the General Electric Company. Only those results are mentioned which are of interest from a theoretical point of view, while the practical side of the question is not considered.

§ 1.

Starting of an Are in Metallic Vapours. Properties of the Cathode.

Metallic vapours, such as vapours of mercury, alkali metals, and some of the heavy metals, which have been investigated, have, even when overheated to a considerable degree, only a very slight conductivity, which in the case of mercury, according to J. J. Thomson (Phil. Mag. [5] vol. xxix. pp. 358 & 441, 1890; see also Strutt, Phil. Mag. [6] vol. iv. p. 596, 1902), is of the same magnitude as that of air. The starting of an are between two mercury electrodes in a well-exhausted tube, by means of moderate voltage, presents, therefore, difficulties which have been overcome by previous investigators in two different ways, both of which were used by Arons.

* Eng. Pat. 211 (1879).

« PreviousContinue »