April 30, 190S] 



JNA TURE 



Goy 



with the pump, &-c. A steel gasometer floating on mercury, 

 and a steel pump working with merCury as a lubricant, had 

 also been constructed for me by Messrs. Brin's Oxygen Co. 

 Sir William Ramsay had placed a large quantity of helium 

 at my disposal. 



Preliminary experiments with oxygen led me to the 

 conclusion that by this method it would be possible to 

 attain to, and measure temperatures far below, the melt- 

 ing point of hydrogen. I may point out that for very low- 

 temperatures the reading of the manometer attached to 

 the thermometer would give a direct measurement of the 

 temperature, as the dead space correction would be very 

 small. IMoRRis W. Th.-wers. 



Indian Institute of Science, Bangalore, March 20. 



The Radio-activity of Ordinary Mttals : the Pene- 

 trating Radiation from the Earth. 

 Is a paper in the Phil. Mag., December, 1907, I 

 described some experiments made by me on the con- 

 ductivity of air confined in metallic cylinders, 60 cm. long 

 and 24 cm. in diameter, whicli were made of lead, of zinc, 

 and of aluminium. With the zinc and aluminium care- 

 fully cleaned, a conductivity was obtained for the enclosed 

 air, which on reduction gave the value 15 for q, the 

 number of ions generated per c.c. per second in the air. 

 With lead cylinders, which were investigated more 

 extensively than those of other metals, the conductivity 

 exhibited wide variations, and values were obtained which 

 ranged from 160 to 23 ions per c.c. per second. 



During the past eight months experiments on the con- 

 ductivity of air confined in metallic vessels have been con- 

 tinued in the Physical Laboratory at Toronto by Mr. C. S. 

 W'right, and he has now obtained under normal conditions 

 with a particular lead cylinder of the dimensions given 

 above, in a series of observations made in a room in the 

 laboratory, a conductivity corresponding to the production 

 of 15-3 ions per c.c. per second. With zinc and aluminium 

 cylinders, the lowest conductivities obtained in this room 

 by him correspond, respectively, to the values 134 and 

 12-5 ions per c.c. per second for q. 



He has also, during this period, conducted a series of 

 experiments on the conductivity of air enclosed in these 

 cylinders in and about Toronto, and has found that the 

 conductivity of the enclosed air varied considerably with 

 the character of the soil and rocks in the neighbourhood 

 of the points of observation. 



In making measurements on the ice above the water of 

 Lake Ontario, the conductivity was found to be very much 

 lower over the surface of the water than at points on the 

 land on either side of the lake at some distance from the 

 shore. In these experiments on the ice the values 8-6, 

 6-0, and 6-55 ions per c.c. per second were found for q 

 with cylinders of lead, zinc, and aluminium respectively, 

 and in a more extended series of observations with the lead 

 cylinder alone, the conductivity was found to be the same 

 over water with depths varying from 2-5 to 10 metres. 

 Measurements were also made on board the steamer 

 Corona during one of her passages over the lake, and 

 values were obtained for q uniformly lower by approxi- 

 mately 6 ions per c.c. per second than those found in the 

 laboratory at Toronto, although the depth of the water 

 at the wharf in Toronto, where the observations in this 

 series were commenced, was not more than 6 or 7 metres, 

 while it was approximately 150 metres in depth at the 

 deepest point on the line of passage. 



Observations made on a sand bar extending out into the 

 lake near Toronto gave a value of 9 ions per c.c. per 

 second for q, and others made on land, at some distance 

 from the shore, at various points and over different soils, 

 gave values ranging from 11.2 to 15 ions per c.c. per 

 second. 



From the investigation it would appear that the water 

 of Lake Ontario, as well as the sand along the shore 

 lino, contains little, if any, radio-active materials, and 

 consequently does not contribute any appreciable propor- 

 tion of the penetrating radiation observed at points on the 

 earth's surface. 



It would appear, too, from the constancy of the observed 



drop in conductivity that the water of the lake completely 

 screens off any radiation coming from the soil or rock 

 beneath it. In order to confirm this view, some experi- 

 ments were made on the absorbing power of the water 

 for the y rays from radium. Thirty milligrams of 

 radium bromide were enclosed in a brass tube with walls 

 about I cm. thick. This tube was laid on the ice, and 

 the ionisation chamber placed 113 cm. above it. With this 

 arrangement it was found that the conductivity added by 

 the radium bromide corresponded to the generation in the 

 air in the chamber of 4485 ions per c.c. per second. A 

 hole was then made in the ice, and the tube was lowered 

 to different depths in the water beneath, the conductivity 

 being measured for each position of the radium. At a 

 distance of half a metre below the surface the conductivity 

 corresponded to the production of 4472 ions per c.c. per 

 second, at i metre to 16-11, at 2 metres to 0-69, and at 

 3 metres to 0-62 ions per c.c. per second. From these 

 numbers it will be seen that a layer of water between 

 2 and 3 metres in thickness sufficed to absorb practically 

 all the radiation issuing from the radium in the tube. 



In view of these experiments and of those of Elster 

 and Geitel, who observed a fall of 28 per cent, in the 

 conductivity of air enclosed in an aluminium cylinder, on 

 taking this cylinder from the surface of the earth to the 

 bottom of a mine surrounded with a wall of rock salt, it 

 would seem that the penetrating radiation observed by a 

 number of investigators at the surface of the earth is more 

 or less local in character, and that, while its existence 

 mav be traceable to active substances present in the soil 

 and rocks, the effective intensity is largely determined by 

 the amount of inactive substances it may have to pass 

 through in order to reach the surface. 



The extremely low values found for q with the cylinders 

 of lead, zinc, and aluminium in the experiments on the 

 ice are interesting on account of their uniformity. They 

 are, as is evident, of the order of magnitude of effects 

 which might easily be accounted for by active impurities 

 in the metals, since differences as large as these values of 

 q may easily be obtained with cylinders made from 

 different samples of almost any metal selected at random. 

 Considering also the difference in the atomic weights of 

 the three substances aluminium, zinc, and lead, and having 

 in mind that radio-activity is a property associated with 

 atomic structure, it would seem tliat if these metals could 

 be obtained entirely free from active impurities, and the 

 conductivity o'f air contained in vessels made from them 

 studied, it would be found, if the observations were carried 

 out under conditions or in places where no ionisation was 

 possible from penetrating radiations arising from external 

 sources, to drop to a very low value, if it did not entirely 

 vanish. 



The experiments described in this note were made with 

 one of Mr. C. T. R. Wilson's latest type of gold-leaf 

 electrometers, which was found, on account of its porta- 

 bility, and of the facility and exactness with which read- 

 ings could be made with it, to be most admirably suited 

 to the purposes of the investigation. 



J. C. McLennan. 

 Physical Laboratory, University of Toronto, 

 March 30. 



The Theory of Dispersion and Spectrum Series. 



On p. 413 of Nature (March j) Prof. Schott attempts 

 to show that there is an irreconcilable inconsistency between 

 Drude's dispersion formula and Balmer's formula for the 

 lines in the hydrogen spectrum. I imagined that some- 

 one who could speak with greater authority than myself 

 would make the obvious reply, but since no such reply has 

 been forthcoming, and the arguments have been re- 

 published in another journal, I venture to ask for space 

 to point out why they appear to me fallacious. 



Prof. Schott's error consists in assuming that the \,, in 

 Drude's formula is the same as the X-, in Balmer's formula. 

 The ^;, in Drude's formula is the wave-length of the light 

 for which the medium shows selective absorption ; that in 

 Balmer's formula is the wave-length of the light emitted 

 by the gas when in a luminous state. An unintelligent 



NO. 2009, VOL. 77] 



