278 CARNEGIE INSTITUTION OF WASHINGTON. 



corpuscle during the ejection of the electron than would be the case in ioniza- 

 tion by a slowly moving corpuscle. It is not possible, in this abstract, to 

 cover many difficulties which may suggest themselves, but the arguments 

 cited serve to show that, in so far as we have at present knowledge of the laws 

 of ionization by very swiftly moving corpuscles, there is theoretical support 

 for the view that as the corpuscular velocity approaches that of light, the 

 efficiency of ionization falls off with great rapidity in the vicinity of that veloc- 

 ity. It may in fact be that for these corpuscles absorption takes place by the 

 complete transfer of the energy of the corpuscle into the 7-ray type during 

 absorption by an atom, in the manner outlined by Bragg, the 7-ray produced 

 being of a quaUty such as to be capable of subsequently ejecting another 

 corpuscle with a speed equal to that of the corpuscle which produced it. 



Note concerning the measurement of ionic density on the top of a tower. W. F. G. Swann. 

 Terr. Mag., vol. 22, 125-127 (Sept. 1917). Washington. 



P. L. Mercanton has recently pubhshed results of measurements of the 

 atmospheric ionization made by him on a parapet near the top of a tower and 

 in the interior of the tower at its base.^ He finds that if n+ and 7i- refer 

 respectively to the positive and negative ionic densities, n+ju- is much 

 larger for the measurement made in the strong electric field near the top of 

 the tower than in the zero electric field in the interior of its base. Thus, for 

 n+ln- as measured at the top of the tower, values 4.33, 2.16, 4.46, and 1.70 

 were obtained on four different occasions, while the corresponding values for 

 the foot of the tower were respectively 1.31, 1.09, 1.24, and 1.19. 



Attention is called to the confirmation afforded by these observations of the 

 view put forward by the author, in a former communication,^ as to the effect 

 of the charge induced on the Ebert ion-counter, by the potential gradient, 

 in modifying the values obtained in ionic-density measurements. The theory 

 of the phenomenon shows that, for a potential gradient of the normal sign, 

 measurements of n+ should be unaffected by the induced charge, while those 

 of n- should be too small. The effect of the negative induced charge on the 

 instrument is to retard the velocity of the approaching negative ions, so that 

 they move slower than the incoming stream of air, Avith a result that fewer 

 ions enter the instrument per second than would enter if there were no induced 

 charge. The positive ions are accelerated by the induced charge, but theory 

 shows that the excess number of ions entering the instrument is just equal to 

 the number which are captured by the outer cylinder of the apparatus, so 

 that the measurements of n+ are unaffected by the induced charge. These 

 conclusions are consequently in complete harmony with P. L. Mercanton's 

 results, since they predict small values of n_ and consequently large values of 

 n+/7i- for places where the potential gradient is high, as, for example, on the 

 top of a tower. 



Reference is fmlher made to some data recently published by A. Gockel con- 

 cerning the effect of a tower on the ionic density in its vicinity.^ In the paper 

 already cited, ^ the problem of the tower is considered, and it appears that, as 

 regards the true negative ionic-density (n-) , the space around the tower may 

 be divided into two regions — a region in which there are no negative ions and a 

 region external to this, in which the ionic-density is normal. For low potential 



^Terr. Mag., vol. 22, pp. 35,-.37, 1917. ^Terr. Mag., vol. 19, pp. 205-218, 1914. 



'Luftelectrische Beobachtungen im schweizerischen Mittelland, im Jura und in den Alpen. 

 Neue Denkschriften der schweizerischen Naturforschenden Gesellschaft, vol. 54, No. 1, pp. 34- 

 35, 1917. 



^Terr. Mag., vol. 19, pp. 205-218, 1914. 



