392 



NATURE 



[February 27, 1896 



interest. I have found that when the Rdntgen rays pass 

 through any substance, they make it for the time being a 

 conductor of electricity, even although the substance is in 

 its normal state a perfect insulator. Thus solid paraffin, 

 paraffin oil, solid sulphur, ebonite, mica, air — all conduct 

 electricity when the Rontgen rays pass through them. 

 This explains the fact observed by Righi, Bergman, and 

 myself, that an electrified plate in air loses its charge 

 when exposed to these rays, whether it be electrified 

 positively or negatively. The air is converted into a con- 

 ductor by these rays, and the charge escapes through it. 

 It is not necessary that a gas should surround the plate, 

 as I have found that the leakage takes place whatever the 

 medium surrounding the plate may be. That this leak- 

 age is due to the condition of the insulator, rather than 

 to that of the plate, is shown by the fact that it occurs 

 when the plane of the electrified discs is parallel to the 

 rays, as well as when it is at right angles to them. The 

 air through which these rays have passed retains traces 

 of conductivity for some little time after the rays have 

 ceased to pass through it ; this can be shown by blowing 

 the air, from a place where the rays are plentiful, against 

 a charged disc placed where there are only a few rays ; 

 the rate of leak from this disc is much increased by the 

 blast. 



With the assistance of Mr. J. A. McClelland, of Trinity 

 College, Cambridge, I have made a large number of 

 measurements of the rate of leak from positively and 

 negatively electrified discs surrounded by air, and have 

 found that the rate of leak in the two cases is almost 

 identical. We have also made a series of measurements 

 of the rate of leak through air at different pressures ; the 

 rate of leak is greater at a high pressure than at a low 

 one, and is over a wide range of pressures approximately 

 proportional to the square root of the pressure. The 

 rate of leak is also greater in air than in hydrogen, being 

 at atmospheric pressure about twice as great in air as it 

 is in hydrogen, while the leakage through carbonic acid 

 gas is faster than that through air. 



The leakage of electricity through non-conductors is, 

 I think, due to a kind of electrolysis, the molecule of the 

 non-conductor being split up, or nearly split up, by the 

 Rontgen rays, which act the part played by the solvent 

 in ordinary electrolytic solutions. If the air through 

 which the rays are passing is ionised, the number of ions 

 would, according to the well-known law of dissociation, 

 be proportional to the square root of the pressure, pro- 

 vided the amount of ionisation is small. Thus the result 

 we obtained for the rate of leak through air at different 

 pressures, indicates that the rate of leak is proportional to 

 the number of ions. 



The view that the air is turned by these rays into an 

 electrolyte, is supported by some experiments made in the 

 Cavendish Laboratory, by Mr. Erskine Murray, on the 

 contact difference of potential between metal plates in 

 air. He finds that when the Rontgen rays are passing 

 through the air in the neighbourhood of the plates, the 

 plates (as far as their potential differences are concerned) 

 behave as if they were connected by an electrolyte. 



Mr. C. J. R. Wilson has investigated, in the Cavendish 

 Laboratory, the effects produced by Rontgen rays on the 

 condensation of clouds caused by the expansion of air, 

 and has found that when the rays pass through the vessel 

 in which the cloud is formed, the cloud is very much 

 denser than when the rays are absent, showing that these 

 rays increase the number of nuclei which act as centres 

 of cloud condensation. The ions with their electrical 

 charges would act as such nuclei, so that this, again, is in 

 favour of the view that these rays turn the air into an 

 electrolyte. These experiments seem to show that these 

 rays exert a powerful disintegrating effect on the 

 molecules of substances through which they pass, and 

 suggest that their use may throw light on some questions 

 of molecular structure. It would be interesting, for 



NO. 1374, VOL. 53] 



example, to find the rate of leak through gases which 

 are reputed to be monatomic, such as mercury 

 vapour, argon, and helium. Again, if air and other gases 

 can be made to'act as electrolytes, we could use a mercury 

 dropping arrangement, similar to that employed by 

 Ostwald to measure the potential difference between 

 metals and liquid electrolytes, to measure, by means of it, 

 the potential difference between metals and various gases 

 through which the Rontgen rays are passing. 



By measuring the rate of leak from a disc charged to a 

 fixed potential, we can compare the efficiency, as pro- 

 ducers of Rontgen rays, of different tubes or of the 

 same tube at different periods. The conclusion I have 

 arrived at, by means of such measurements, is that bulbs 

 generally improve for some time after they are sealed oft 

 from the pump, and attain a maximum efficiency, after 

 which they begin to deteriorate. 



Some measurements of the absorption of Rontgen 

 rays by various thicknesses of metal, have led me to 

 the conclusion that the Rontgen rays are not all of the 

 same kind. The experiments were made by measuring 

 the change of the rate of leak from an electrified disc, 

 produced by changing the number of sheets of tinfoil 

 interposed between the disc and the phosphorescent tube. 

 When a small number of sheets of tinfoil were interposed, 

 the addition of another sheet of tinfoil produced a very 

 considerable diminution in the rate of leak ; when, how- 

 ever, the phosphorescent bulb was a very good one, a 

 considerable leakage remained when the number of 

 sheets of tinfoil was considerable, and this "residual 

 leakage " diminished but slowly as the number of sheets 

 of tinfoil was increased. This seems to indicate that 

 while there are some rays which are rapidly absorbed by 

 the tinfoil, there are others which can pass through it 

 with comparative facility. It is only when using one or 

 two of the most efficient bulbs that I have remarked this 

 " residual leakage," but with these bulbs when in their 

 most efficient state it was very marked. 



J. J. Thomson. 



THE DEEPEST SOUNDING YET KNOWN. 

 TN Nature (vol. Hi. p. 550, October 3, 1895), I "^en- 

 -^ tioned that H.M.S. Penguin^ Commander A. F. 

 Balfour, R.N., had found in the Pacific Ocean deeper 

 water than any yet known in lat 23"4o S., long. i75'ioW., 

 but had failed to reach the bottom owing to breakage of 

 the wire at 4900 fathoms. 



Captain Balfour has been enabled to try again, and I 

 have just received the announcement of his obtaining 

 three satisfactory soundings of over 5000 fathoms. 



The deepest trustworthy sounding heretofore known is 

 4655 fathoms near Japan, obtained by U.S. S. Tuscarora 

 in 1874. 



The deepest of tbe PejtguMs casts is 5155 fathoms, or 

 500 fathoms (3000 feet) deeper ; but it is especially 

 remarkable that the three casts now obtained are not in 

 the same hollow, but are separated by areas of consider- 

 ably less water, the two extreme soundings being 450 

 miles apart. 



The usual abysmal red clay was brought up by the 

 sounding-tube on two occasions ; on the third the wire 

 broke. 



Mr. V. Thorpe, surgeon of the Penguin, reports that 

 a microscopic examination of the specimen from 5147 

 fms. shows that the remains of siliceous organisms are 

 almost, if not entirely, absent. The mineral particles are 

 in a minute state of disintegration, and consist of 

 exceedingly fine flocculent matter, mixed with pumice 

 and other glossy volcanic products, green crystals of 

 augite and reddish crystals of pelagonite. 



These deep hollows furnish fresh evidence to the 

 observed fact that all the extreme depths in the ocean 



