150 



OCEAN ATMOSPHERIC-ELECTRIC RESULTS 



of the fog or mist layer was somewhat greater than that 

 of the preceding haze layer, namely 1.3 km, untU the wind 

 freshened and changed to southwest on the fourteenth, 

 at which time the layer became very thin, about 0.1 km. 

 For the latter period a thin layer would be expected, 

 since the current density was so nearly normal; the fog 

 or mist increased the resistivity in this thin layer by 

 about 50 per cent, but the increased columnar resistance 

 in the layer added only about 3 per cent to the total col- 

 umnar resistance, if the assumption is valid that the 

 total potential E was from day to day the same as it 

 would have been had not fog or mist been present. 



The second period of interest is that in the Atlantic 

 Ocean between August 10 and 25, 1928. During this peri- 

 od the conductivity was about half normal value, and the 

 potential-gradient slightly lower than the average for all 

 undisturbed values tn the Atlantic, so that computed val- 

 ues of air -earth current density for the period had an 

 average value of approximately half of that obtained for 

 the remainder of the Atlantic data. The ship remained 

 between 311° and 322° east longitude during the thirteen- 

 day period under discussion, but sailed southward 1700 

 miles from 40° north down to 15° north. The daily set 

 of observations was generally made about 18h GMT. 

 Table 4 shows the values of the atmospheric-electric ele- 

 ments for August 10 to 25, and again the last column In 

 the table contains values of current density derived on 

 the basis that the lower layer in the atmosphere, in which 

 the conductivity was about half normal value, had a defi- 

 nite vertical extent. 



Table 4. Period of low values of air -earth current 



density in the north central Atlantic, August 10 to 



15, 1928, and derived equivalent fair-weather 



values based on assumed height of affected 



region of the atmosphere 



Utilizing equation (11), as before, preliminary in- 

 spection of the data indicated that the layer having higher 

 than normal resistivity was quite thick. This was appar- 

 ent from the fact that, the potential-gradient being very 

 little different from normal, the reduced air-earth cur- 

 rent density would be indicative of a significant change in 

 columnar resistance of the atmosphere. Accordingly, 

 the value of h^ was chosen such that the second term of 

 equation (11) became zero. This is 



{ Pnhu/Rn) = 



(13) 



With the columnar resistance R taken as 1.11 x lO^esu, 

 as before, and the values of Pn in this case as 0.461 x 

 104 esu (the average value of total conductivity over the 

 Atlantic was found to be 2.17 x 10-4 esu for least-dis- 

 turbed days on cruise VII) the value of hu is found to be 



2.7 km. In effect, the present procedure means that all 

 the columnar resistance is coi^lned to a height of 2.7 

 km, which is not true, but it may be within a reasonably 

 good approximation of beii^ true. The values of in 

 foimd on this basis average 8.0 x 10-7 esu, a value close 

 to the average of 8.7 x 10-7 esu found for all least-dis- 

 turbed days over the Atlantic. 



In this case, whatever may be the nature of the par- 

 ticles or nuclei responsible for the low air-earth cur- 

 rent density, the region containing the particles ^pears 

 to have extended vertically to a considerable distance. 



The third and last group of material containing im- 

 usual values of air -earth current density was obtained 

 between September 9 and 20, 1929, after the ship left San 

 Francisco on the voyage to Honolulu. It will be recalled 

 from an earlier paper In this volume that a few days after 

 leaving San Francisco the conductivity decreased in the 

 course of one day to about one-tenth its normal high value, 

 and subsequent recovery to the original value required a 

 period of ten days. The day on which the drop occurred 

 was September 9, 1929, and the present discussion will 

 deal only with that day. The hourly mean values of con- 

 ductivity for September 9 (negative conductivity recorded 

 on this date) have been tabulated on pages 124 and 125 in 

 the table in section Vm, and corresponding hourly mean 

 values of potential-gradient on pages 120 and 121 in sec- 

 tion vn. For convenient reference the data have been 

 reproduced in table 5 below, where it will be noted that 

 five groups have been arranged, the first (5h to llhGMT) 

 representing least-disturbed or normal conditions, and 

 the remaining four showing progressive changes in the 

 various elements tabulated. 



When the conductivity and potential-gradient data are 

 examined together, it is clear that as the conductivity de- 

 creased through the Greenwich day, the potential- gradient 

 increased to much higher values than would be expected 

 on the basis of the normal universal diurnal variation. 

 In group (1) in table 5 the potential-gradient averaged 

 123 volts per meter, representing least-disturbed con- 

 ditions in the early hours of the Greenwich day, and the 

 normal increase to maximum at 18 to 20 hours GMT 

 might have been expected to give a value for those hours 

 of perhaps 180 volts per meter instead of the 360 volts 

 actually recorded. Air-earth current density also wotild 

 have increased through the Greenwich day, had the day 

 been normal, from 10.4 x 10-7 esu as found for the 

 period 5 to 11 hours GMT, to perhaps 13 x 10-7 esu or 

 more at 18 to 20 hours GMT. In table 5, under the col- 

 umn in, assumed values of "normal" current density 

 are given for groups (2) to (5) to represent the diurnal 

 change in this element with universal time. These val- 

 ues were arrived at with the help of the lowermost curve 

 in figure 14, which gives the diurnal variation, on Green- 

 wich time, of the air-earth current density for October 

 and November, 1929. On that curve the average current 

 density for the six-hour period from 5 to 11 hours, is 



9.8 X 10-7 esu, which is 0.6 x 10-7 esu lower than the 

 value found for group (1) in table 5. For groups (2) to 

 (5) comparable values from the curve are 11.0, 12.1, 

 11.7, and 10.0 x 10-7 esu, respectively, and when these 

 are adjusted upward by the difference of 0.6 x 10-7 esu 

 found between group (1) and the curve, the best available 



