Atmospheric-Electric Observations, 1907-08 365 



Rain-water, caught as it fell and immediately evaporated to diyness, showed no sign 

 of radioactivity. The electroscope readings were always difficult, and not of sufficient 

 accuracy to detect extremely small effects. The electroscope was always placed so as to 

 allow the leaves to swing in a plane parallel to the length of the ship, so as to eliminate the 

 effect of rolling as far as possible, but the leaves were never quiet and their mean position 

 had to be estimated. 



The only really satisfactory instrument for regular use on board ship was the Gerdien 

 apparatus for determining the specific conductivity of the air. An Ebert ion-counter was 

 also included in the outfit, but its leakage was too great and the time necessary for a single 

 determination too long, so no results were obtained with it. 



The Gerdien conductivity apparatus was the same as used by J. E. Burbank in his 

 work in Labrador during the eclipse of 1905.^ A uniform current of air is drawn by means 

 of a fan through a cyUndrical condenser, the inner cylinder of which is connected with the 

 leaves of an aluminum-leaf electroscope. The outer cylinder is 16 cm. in diameter and 35 

 cm. long, while the inner cylinder is 1.4 cm. in diameter and 24 cm. long. The capacity 

 is 12.9 cm. The inner cylinder being charged to a known potential, read on the electro- 

 scope, an is drawn through for a measured interval of time, usually 5 minutes. The ions 

 of opposite sign to the charge on the cylinder will be attracted to it from the air passing by, 

 and a certain portion of the charge will thus be dissipated. Only those ions will reach the 

 inner cylinder which have sufficient velocity to carry them across the intervening space 

 before they are carried by. The number of ions reaching the inner cylinder is practically 

 independent of the velocity of the air-current so long as it is sufficient to prevent saturation 

 currents from being established, and it is only necessary to insure that the velocity does not 

 fall below a certain minimum value. Knowing the capacity and dimensions of the instru- 

 ment and the time during which the air-current has been passing, the specific conductivity 

 of the air can be computed from the potential of the inner cylinder at the beginning and the 

 end of the exposure. 



The instioiment was at first mounted on a ship gimbal-stand, which was placed on top 

 of the forecastle, under the observing bridge for magnetic observations, and for one-half 

 of the voyage the observations were made at that place. The location was, however, not 

 satisfactory on account of the neighborhood of the galley smokestack, smoke from which 

 often reached the instrument during calms or while saiUng by the wind. Accordingly, on 

 the cruise between New Zealand and Peru (on February 3, 1908) the gimbal stand was 

 moved to the main deck, just forward of the main hatch and still under the bridge. Here 

 there was no further trouble from smoke. 



The measurements of the specific conductivities \+ and X_ for positive and negative 

 ions gave as means^ 



X+ = 1.603 X 10-* E.s.u. (from 258 observations) 

 X_ = 1.433 X 10-* E.s.u. (from 260 observations) 



^ = 1.12 



A_ 



The barometric pressm-e apparently affects the conductivity, as the mean values for 

 \+ and X_ for 34 days with the pressure below 762.0 mm. are 1.61X10"* e.s.u. and 

 1.46X10"'* E.s.u., respectively, while for 24 days with the pressure 762.0 mm. or above, the 

 mean values forX^ and X_ are 1.52X10"* E.s.u. and 1.31X10"* e.s.u. respectively. High 

 barometer apparently causes a decrease of conductivity. This may, however, be due to 

 the fact that the low barometric readings were nearly all within the tropics, while the high 



'See Terr. Mag., vol. 12, pp. 97-104, 1907, for description of instrument. 

 ^The original report contains also the complete tables of the individual values. 



