320 CARNEGIE INSTITUTION OF WASHINGTON. 



insulator was designed for supporting the ionium collectors used in the poten- 

 tial-gradient observations. This insulator was provided as a precaution 

 against insulation troubles due to rain, extreme humidity, and the activities 

 of small insects, and proved to be entirely satisfactory. The apparatus used 

 for potential-gradient observations was standardized by the method described 

 by Simpson and Wright.^ 



The actual observations were made by Mr. Thomson, assisted by Mr. 

 Antonio Lima. On May 29 the observations were made, as nearly con- 

 tinuously as was consistent with adequate control observations, from sunrise 

 until middle of the afternoon. On all other days from May 24 to June 2, 

 inclusive, regular observations for comparison purposes were made hourly, 

 for periods of 20 minutes each, throughout the forenoons. Average meteoro- 

 logical and atmospheric-electrical conditions prevailed on May 29, except 

 that the sky was overclouded just before and during the first half of the eclipse. 

 If this is taken into account the general results may be given as follows : 



(a) The potential gradient showed a well-formed minimum which began 

 with totality and continued until about 20 minutes after totality. The 

 values observed during this period were about 20 per cent lower than a mean 

 value derived from the two 20-minute periods which immediately preceded 

 and followed it, respectively, and also lower, in about the same proportion, 

 than the mean value obtained from the observations for the corresponding 

 periods on comparison days. 



(h) During the above period of minimum value the fluctuations of the 

 potential-gradient were very much smaller than during the two similar periods 

 referred to under (o) . 



(c) If the values of the positive and negative conductivities observed 

 during a period of about 15 minutes just following totality are compared with 

 the mean obtained from two periods of equal length just before and shortly 

 after the one under consideration, it appears that there was, during this period, 

 an increase of the order of 20 per cent for both. A similar remark applies 

 if the comparison is based on the average value for the given time of day as 

 obtained from May 24 to June 2. 



(d) The air-earth current-density, as computed from the potential-gradient 

 and total conductivity, showed more constancy and adherence to the normal 

 station value during the period in which the obscuration of the sun was at 

 least 50 per cent than for any equal period of the day of the eclipse. 



From the above it appears that the results for May 29, 1919, at the Sobral 

 station are in general agreement with those obtained at Lakin, Kansas, during 

 the eclipse of June 8, 1918,^ notwithstanding the marked differences between 

 the two stations as regards latitude, elevation, general topography, and dis- 

 tance from sea. 



A cruise on the brigantine Carnegie. N. Meisenhelter. National Marine, vol. 15, No. 6, 



10-18 (June 1920). 

 Earthquake of April 30, 1919, as recorded on the magnetogram at the VVatheroo Magnetic 

 Observatory. W. F. Wallis. Terr. Mag., vol. 24, 175 (December 1919). 



This note gives the particulars of the earthquake effects registered on the 

 magnetograph record for April 30, 1919, at the Watheroo Magnetic Observa- 

 tory. This record is doubtless that of the earthquake and tidal wave which 

 swept Pan-Gai, a town in one of the islands of the Tonga group approximately 

 4,640 miles east-northeast of the Observatory. 



iProc. R. Soc. A., vol. H5, p. 182, 1911. 



■^Terr. Map;., vol. 24, p. 96, .Juno 1919; Annual Report of the Director of the Department of 

 Terrestrial Magnetism for 1919, p. 307. 



