February ii, 1909 J 



NA TURE 



443 



accurate estimate of any of the charges has not been 

 obtained owing to the difticulties of experimenting with 

 the clouds. As no trustworthy ' independent estimates have 

 been made of h, the vaUie of the product nxc for gaseous 

 ions can only be obtained by this method within wide 

 limits ditTering by a factor of lo or 20. It cannot, there- 

 fore, be maintained that the direct determination of e in 

 gases leads to any trustworthy information as to the 

 simple relations that hold between the charges on the ions. 



A more accurate comparison of the charges on the 

 various kinds of ions can be obtained from determinations 

 of the rate of diffusion of ions in gases and the velocity 

 under an electric force. With this object in view, the 

 rates of diffusion produced by various methods in gases 

 were determined, and it was shown that the value of 

 «Xt' for negative ions in gases agreed within 10 per cent, 

 or ij per cent, with the value for monovalent ions in 

 liquids, and the value for positive ions in gases was some- 

 what larger (J. S. Townsend, " Diffusion of Ions in 

 Gases," Phil. Trans., vol. c.\ciii., "1899, and vol. cxcv., 

 igoo). The probable error in the numbers obtained is 

 about 10 per cent, or 12 per cent., so that it is desirable 

 to know more definitely if all these charges are exact 

 multiples of the same atomic quantity, as it is a question 

 of fundamental importance. 



The problem of the determination of n'x.e for gases has 

 been again undertaken, and a simple experiment has been 

 devised whereby the exact value of nxe can be immediately 

 deduced from the ratio of the charges acquired by two 

 conductors under special conditions. The method is ex- 

 plained in a paper in the Proceedings of the Royal Society, 

 vol. Ixxx., January, 1908, and two papers recently com- 

 municated (November, igo8) contain further experiments 

 by the present writer on ions produced by Rbntgen rays, 

 and an investigation by Mr. Haselfoot of the ions produced 

 by radio-active substances. 



The principle of the method consists in finding the extent 

 to which a uniform stream of ions having a circular 

 cross-sectional area, S, opens out as the ions travel a given 

 distance under a known electromotive force. For this 

 purpose three plates, k, B, and C, are arranged parallel 

 to each other, the middle plate, B, and the lower plate, C, 

 having each a circular aperture cut through its centre. A 

 disc, D, is fi.xed in the aperture of the plate C, so that the 

 surfaces of the disc and surrounding plate are in the same 

 plane, the disc being a little smaller than aperture in 

 order to insulate it from the plate. The area of the hole 

 S in the middle plate B is equal to the area of the disc 

 j*hts half the air-gap between the disc and the plate C. 

 The plates .A and B are connected to suitable numbers of 

 accumulators so as to maintain the same uniform field 

 above and below the middle plate B. The plate C and 

 disc D are insulated, and each maintained at zero poten- 

 tial by a special form of induction balance, which gives 

 the charges acquired simultaneously by the disc D and 

 plate C. The gas in the space between A and B is ionised 

 by Rontgen rays or by radio-active substances, and a 

 uniform stream of ions passes through the aperture in 

 the middle plate. The ions travel to the lower plate under 

 Ihe uniform electric field, and the stream opens out by 

 diffusion, so that some of the ions g, arrive on the disc 

 D, and the rest c\., arrive on the plate C. The ratio (],ftj; 

 is found accurately bv means of the induction balance, and 

 the value of 11 x^ may be obtained from the ratio. The 

 equation connecting nxc and the ratio njn^ is somewhat 

 complicated, and it would be impossible to explain in a 

 short space how the connection between these quantities 

 is found, but it may be stated that a complete solution of 

 the problem can be obtained in a series of Bessel's 

 functions. 



The experiments have been made with different forces 

 and pressures, and it has been found that the value of 

 II Xe for negative ions is in all cases within 3 per cent, or 

 d per cent, of the value 1-23x10'°; under conditions where 

 the greatest accuracv can be obtained the results are in 

 rloser agreement with this number. 



For positive ions the value of nxe depends on the nature 



1 Prof. Perrin has recently announced a new method of de'ermininc it, 

 which gives trustworthy results. The nuinher « comes to ,^y lo'!' and corre- 

 soonds to an atomic charge 4'i X io-»'. (Jean Perrin, Cotiiftcs remlus, 

 October 5, ioc8). 



NO. 2050, VOL. 79] 



of the radiation. With non-penetrating secondary rays 

 from a polished metal surface the value obtained was 

 1-26x10'", and for penetrating rays from a tarnished 

 surface, or a "surface covered with a thin layer of vaseline, 

 larger values were obtained, the greatest being 2-4x10"'. 



Uhus the negative ions have always a charge which is 

 exactly equal to the charge on a monovalent ion in a 

 liquid' electrolyte, and the positive ions have either a 

 single or a double charge, the number of either kind in 

 a conducting gas depending on the nature of the radiation. 



The values of 11 xe for positive and negative ions pro- 

 duced by the a. and /3 rays from radio-active substances 

 are both approximately 1-23x10'". 



In addition to the above results, a notable effect of 

 small traces of moisture on the motion of negative ions 

 was observed. When the gas is very dry the negative 

 ions move as if they were very small particles, but when 

 a small amount of moisture is admitted the mass of the 

 negative ion is greatly increased, and obeys the same 

 laws of diffusion as the positive ions. The motion of the 

 positive ions under similar conditions is not affected by 

 the dryness of the gas. John S. Townsend. 



METEOROLOGICAL CHARTS OF THE 

 INDIAN OCEAN.' 



THE Indian Ocean is claiming at the present time a 

 large share of the attention of meteorological offices. 

 Recent issues of Nature have contained notices of meteor- 

 ological charts for this area issued by the Meteorological 

 Department of the Government of India and by the 

 Meteorological Institute of the Netherlands (N.4TURE, vol. 

 Ixxviii., pp. 169, 487). The present charts are prepared by 

 the Deutsche Seewarte. In area they exceed considerably 

 those referred to above, for they embrace the region 

 between latitudes 30° N. and 50° S., and longitudes 18° E. 

 (Cape Town) and 158° E. The Australian waters and the 

 eastern margin of the Pacific Ocean are thus included, 

 while special inset charts extend the area northwards to 

 include the Yellow Sea and the Sea of Japan. To deal 

 effectively with the results, a scale of approximately 6 mm. 

 to one degree at the equator has been selected, and in 

 consequence an inconveniently large size of page, viz. 

 36 inches by 27 inches, has had to be adopted. 



The preparation of the results has occupied five years. 

 The meteorological information has been abstracted mainly 

 from the log-books of German vessels, but we are glad 

 to note that, in addition, use has been made of all avail- 

 able published information. The arrangement of the data 

 on the charts, of which there is one for each month, js 

 similar to that adopted on the charts for the Atlantic 

 Ocean issued by the Seewarte. Conspicuous blue wind 

 roses show for each square of 5° the percentage frequency 

 of calms and of winds from each of sixteen directions. 

 The mean wind force for each direction, on the Beaufort 

 scale, is indicated by the number of barbs on the wind 

 arrows. Small but distinct black arrows give the direc- 

 tions of surface currents, with the average and the maxi- 

 mum observed displacement in nautical miles per day. 

 Special attention has been devoted to a critical examina- 

 tion of the current data, and several interesting articles 

 on the subject appear on the backs of the charts. A state- 

 ment of the number of observations on which each wind 

 and current arrow is based would have been welcomed by 

 students. 



In addition, each chart gives the tracks for steam and 

 sailing vessels, the normal paths of hurricanes, the fre- 

 quency of fog and Ice, and the lines of equal magnetic 

 declination. The region of easterly variation is dis- 

 tinguished by a special tint. The text printed over the 

 land areas gives, in addition to the necessary explanations, 

 a brief summary of the weather conditions of each month, 

 wilh special reference lo the frequency of hurricanes. 



On the back of each chart we find four smaller maps, 

 giving the annual change of magnetic variation, the 

 average air temperature over sea and land, the average 

 temperature of the surface water, and the average baro- 

 metric pressure. In connection with the latter, we miss 



1 Deutsche Seeware, Monatskarten fiir den indischen Ozean. (Hamburg 

 Eckardt und Messtorff, n.d.) 



