REPORT ON ATMOSPHERIC ELECTRIC WORK 



45 



occurred with this instrument. It has been possible, 

 with a new type of hood over the apparatus, to observe 

 during periods of light sprinkling rain without any bad 

 effect on the insulation of the apparatus. 



Penetrating Radiation Apparatus 1. — The sodium in 

 the drying tubes was renewed on October 16. The new 

 tubes supplied at San Francisco were foujid to fit; so we 

 now have an ample supply on board. On October 17 ob- 

 servations with this instrument and penetrating radia- 

 tion apparatus 5503 were made, using radium as an 

 ionizing constant. The results appear quite satisfactory 

 but it is planned to repeat them at an early date. 



Penetrating Radii^ion Apparatus 5503. — The diurnal - 

 variation curve of October 5 and 6 showed irregularities 

 which obviously were instrumental and were bronght about 

 in the following way. The electrometer was charged up 

 every two hours, and, owing to the desirability of mak- 

 ing the mean times of observation one hour apart, the 

 initial reading of the first, third, fifth, and subsequent 

 odd-numbered hours was made one minute after re- 

 charging. This arrangement was adopted because the 

 electrometer had been charged for about two hours be- 

 fore the beginning of the diurnal-variation run, and its 

 performance during that period indicated that initial 

 readings could be taken very shortly after recharging. 

 This, however, later was found not to be justified. In 

 order to determine the minimum time required for nor- 

 mal values to be obtained, a series of observations was 

 made on October 8. The fibers were charged and read- 

 ings were made after one, three, five, etc., minutes had 

 elapsed. After twelve of these readings the instrument 

 was left for one hour and the final reading then made. 

 It was found that the "charging effect" persisted for 

 about five minutes rather than only one minute, after 

 which the values of R were consistent, within the limits 

 of accuracy of the instrument. The hourly routine of the 

 diurnal -variation observations therefore has been 

 amended, as follows: 



h m 



Performance 



00 05 First reading of 5503 



00 15 First reading of penetrating radiation appa- 

 ratus no. 1 



Leak test of ion counter 

 00 20 Begin ion count 



Nuclei count and meteorological observations 

 00 40 End ion count. Start leak test of ion counter 

 00 45 Last reading of penetrating radiation appa- 

 ratus no. 1 

 00 55 Last reading of 5503 



Recharge 5503 immediately 



Above routine repeated each hour 



Two of the diurnal-variation runs (October 13 and 14 and 

 21 and 22) show very good agreement between instru- 

 ments 1 and 5503. Both curves show a pronounced double 

 maximum. The curves on the other two occasions dur- 

 ing the period covered by this report are more erratic. 



Radioactive Content Apparatus 4. --Difficulties with 

 the electrometer used with this apparatus on October 3 

 and 4, led to exhaustive tests and finally to the exchange 

 of this Instrument for the spare, no. 15, which has flat 

 plates instead of the knife-edge type of plates of no. 5. 

 There was no further difficulty. Frequent attention was 

 necessary to the collecting apparatus. The values ob- 

 tained on October 8 and 10 were high; since that date 

 low values have been recorded and a test observation 

 made just before entering Pago Pago harbor on Novem- 

 ber 18 showed a decided increase, tending to show that 

 the apparatus had been functioning properly. On two 

 occasions collection had to be made without the upper 

 earthing cap, but after tightening the friction clamp at 

 the top, a high potential could be maintained and so the 

 cap was used on subsequent observations. 



Nuclei Counter 5. --Regular observations have been 

 made and, in general, low values have prevailed. 



Batteries. --The Burgess "B" batteries continue to 

 give good service. 



