July 23, 1896J 



NATURE 



;Si 



elusion is that the intensities of all the lines were nearly propor- 

 tional to the mass of air traversed. The lines 6868 (.8), 6517, 

 6278 (a), and 5800 (8) were practically of the same intensity, for 

 equal masses of air, at all the three observinp; stations, showing 

 that the presence of water-vapour had little or no influence on 

 them, and indicating that their origin was most probably atmo- 

 spheric oxygen. The rain-band line (5943), however, has,, for 

 equal masses of air, a much less (about one-third) intensity at 

 Etna than at the other two station.s. The mean vapour tensions at 

 the three places, Etna, Nicolosi, and Catania, were as 3 : 7 : 10, .so 

 that this line 5943 is evidently due to aqueous vapour. The fact 

 that when the observations are plotted the curves pass through 

 or near the origin, indicates that the atmospheric oxygen and 

 water-vapour are the sole causes of these telluric lines. 



E.XI'LANATION OF SOLAR PHIiNOME.SA. — In the June 

 number of the --/j/rtf Physkal Journal, J. Fenyi discusses several 

 new explanations of the various features of the solar surface, 

 emphasising several physical facts, hitherto neglected, the con- 

 sideration of which simplify the conception of the causes of the 

 solar phenomena. He assumes that the prominences are masses 

 of real matter in violent motion, and also that they are ejected 

 intoy;rt' space. The crucial point of his argument is that when 

 a mass of hydrogen, say, is ]>rojected from the photosphere, 

 and has passed through the chromosphere into free space, it is 

 not diffused immediately, but takes a certain time, termed the 

 expansion interval, which varies directly as the diameter of 

 the mas.s, and is inversely proportional to the square root of 

 the absolute temperature. By following out in detail the 

 phenomena of eruptive prominences, he explains them all on 

 this view, especially their unusual brightness and rapid dis- 

 solution. The while prominences he accounts for as being the 

 expanded gaseous portions of former ordinary prominences, 

 rendered visible by reflected sunlight. The corona he regards 

 as being due to more distant masses of these gaseous materials, 

 primarily ejected as prominences, the enormous length of some 

 coronal streamers being no difficulty if they are admitted to be 

 projected in free space. Facule »lll then be produced by these 

 gaseous matters falling down on the photosphere again, their 

 superior brightness being due to the heat generated during their 

 fall, together with the actual radiation received from the sun 

 itself. Their prevalence in sun-spot zones is explained if they 

 are the consequence of eruptive prominences, which themselves 

 favour these zones. This dispenses with the view that facuU^ 

 are projected prominences, and regards them as the result of 

 prominence action. The much-disputed question of the reason 

 of distorted spectral lines in prominences is greatly simplified by 

 this explanation. If a mass of ascending gas as a prominence 

 encounters a descending mass from a previous eruption, the 

 resultant motion will in general be tangential to the solar sur- 

 face, and will be capable of producing the enormous velocities 

 in the line of sight which have been measured in prominence 

 spectra, and which could not be explained as being the result of 

 mere explosions from the photosphere. 



NEW FORM OF APPARATUS FOR THE 



PRODUCTION OF RONTGEN RA YS. 

 COME time in the month of March this year, after working 

 *~^ with various forms of tubes, it occurred to the writer to 

 abolish the glass vessel by converting the ordinary concave 

 kathode into a nearly complete sphere, with the platinum anode 

 at its centre. A simple experiment with a Jackson bulb proved 

 that the rays from the anode could pass through the material 

 of the kathode as they would through a similar piece of un- 

 electrified aluminium placed outside the bulb. Hence it became 

 fairly evident at the outset that the proposed plan would work 

 to some extent. 



Under the guidance of Prof. Lodge, and in his research 

 laborator)', experiments were commenced. The first arrange- 

 ment was a simple one. The sphere was made in two halves, 

 one half of copper and the other of aluminium. The two halves 

 were joined together with marine glue only. The anode was 

 held in position by ebonite fixed in the co]5per hemisphere. A 

 section of this simple arrangement is shown in Fig. i. The 

 section is drawn to scale, the diameter of the sphere being 2 

 inches. This early apparatus .showed signs of success, and it 

 was decided to invest in a larger sphere — one of 3! inches in 

 diameter. The joints were now made much more carefully, 



NO. 1395, VOL. 54] 



and the apparatus so designed that it could be fitted together 

 or taken to pieces in half an hour's time. The hemispheres of 

 copper and aluminium were soldered together, but the joints 

 (A and B, Fig. i) were made by compressing indiarubber washers 

 by means of suitably made screws. With this convenient appa- 

 ratus the behaviour of various sizes and shapes of anodes was 

 observed. In all the experiments a small thick plate of platinum, 

 having a plane surface of about \ square inch, was reserved for 

 that portion of the anode which received the kathode rays ; 

 the remainder of the anode was sometimes of aluminium and 

 sometimes of copper. The various forms tried are shown in 

 Figs. 2 to 8. 



Ft. 



In Fig. 2 we have the simplest possible anode — the platinum 

 plate alone. It is the same arrangement as that of Fig. 1, the 

 only difference being that of dimensions. This form possessed 

 an enormous resistance, so that only with low vacua could a 

 current be made to pass through. For this reason the be- 

 haviour of this form was unsteady and its periods of activity 

 very short. With higher vacua and greater potentials, no doubt 

 this form would be more successful. Another form tried was 

 that shown in Fig. 3. The anode here was very considerably 

 enlarged by placing a circular plate of metal just behind the 

 platinum, at a place where no kathode rays could fall on it. 

 By this means the area of the anode surface was increased 

 sixteen-fold approximately. The resistance was thereby much 

 reduced, and it became possible to work at higher vacua. This 

 form gave a more powerful and a considerably more uniform 

 radiation than that of its predecessor. 



The next step was to increase still further the area of the 

 anode (see Fig. 4). The anode now nearly filled the sphere. 

 The result, however, was not so good, tending to show that the 

 best size of anode is something less than Fig. 4, and greater than 

 Fig. 2 ; but Prof. Lodge thinks that this is a question of the 

 particular vacuum employed. Another differently-shaped anode 

 was next tried. This was formed of a metallic hemisphere with 

 a flat plate in front of it (see Fig. 5I. The idea was to get all, 

 or nearly all, of the electric discharge, and so possibly most of 

 the kathode radiation also, to take place between the outer 

 aluminium hemisphere and the anode. The idea probably is a 

 crooked one ; anyhow, this form proved less successful than 

 others. The plate was next removed, and the hemisphere \vas 

 replaced by a larger one, as in Fig. 6. For some unknown 

 reason this form gave no radiation whatever, although the 



