STUDIES IN ATMOSPHERIC ELECTRICITY 



147 



interest. These are found on November 21 to 22, 1928, 

 and May 17 to 18, 1929, and the value in each case is 6.2 

 X 10-'' esu. In these two cases the values of conductiv- 

 ity are typical of those obtained under least-disturbed 

 conditions, but the potential-gradient values are lower 

 than average. Now, storm clouds of proper polarity at 

 appropriate distances might be thought of as causing the 

 low potential-gradients on these occasions, but the ab- 

 stract of the log in section IV does not indicate that such 

 clouds were present. Another explanation for the low 

 values of potential-gradient may be found in the change 

 in this element with latitude. Gish (4) found a conspicu- 

 ous change with latitude in his study of the potential-gra- 

 dient data from the earlier cruises of the Carnegie . Al- 

 though he did not find a comparable change in air -earth 

 current density with latitude from the data for earlier 

 cruises, the data of the present cruise, which show that 

 under least-disturbed conditions the conductivity is es- 

 sentially constant, require that the current density 

 change with latitude in accordance with the change in 

 potential-gradient for different latitudes. 



When the current-density values in table 2 are plot- 

 ted against latitude, an increase with latitude is indicat- 

 ed, but the two low values under discussion still appear 

 low for the latitudes at which they were obtained (11° 

 south and 16° north). There remain two ways of account- 

 ing for the low values of current density on these occa- 

 sions. Either an unusually low value of the total potential 

 E may have existed between the earth and the upper con- 

 ducting region, or there may have been an unusually high 

 value of columnar resistance R. Since, however, the 

 conductivity appeared undisturbed, the value of R proba- 

 bly was not unusually high, and in that case the low po- 

 tential-gradient and low current density on these occa- 

 sions was caused by a low value of E. That the value of 

 E might on one day be low and the next day high, say half 

 again as great or even more, does not appear unreason- 

 able. 



From the preceding discussion it may be seen that 

 the air-earth current density may have high or low val- 

 ues through the day or from day to day from various 

 causes. As one cause, some disturbing element may en- 

 ter the lower layer of the atmosphere as evidenced by 

 smoke, dust, fog, mist, or haze, to lower the current 

 density to a smaller value than would prevail if this ele- 

 ment were absent. As a second cause, observing in mid- 

 dle or low latitudes would yield smaller values of current 

 density than would measurements in high latitudes. As a 

 third cause, the total potential E might vary from one 

 occasion to another to produce higher or lower values 

 from one time to the next. In the discussion to follow, 

 account will be taken of these various causes in a study 

 of disturbed periods encountered during cruise VII. The 

 disturbed periods will be compared with so-called least- 

 disturbed periods, and for the latter periods it will be 

 assumed that the columnar resistance is constant and 

 has the value given by Gish. Departures from this value 

 in periods of disturbance will be evidenced by unusually 

 low values of conductivity, the degree of "lowness" 

 being indicated by comparison with a value derived from 

 measurements of this element obtained under least-dis- 

 turbed conditions and accepted as essentially constant. 

 There will also prevail, in the disturbed periods, low 

 values of current density, and the degree of "lowness" 

 in this element will be indicated by comparison with a val- 

 ue representing least-disturbed conditions after giving prop- 

 er weight to the effect of latitude and of variations in E. 



The first of the three disturbed periods encountered 

 on cruise VTI to be considered will be that of June 2 to 

 July 21, 1929. Throughout the thirty-three days at sea 

 during this period the atmosphere was hazy, foggy, or 

 misty, or a fine drizzling rain was falling, as recorded 

 in the abstract of the log in section IV. The periods 

 Jime 2 to June 7, before arriving at Yokohama, and June 

 26 to July 2, just after leaving that port, were character- 

 ized by haze. The period June 2 to 7 was also marked by 

 two typhoons, the first reaching the vicinity of the Car- 

 negie on June 2 and the second on June 6. That the haze 

 encountered before arriving and after leaving Yokohama 

 is tj^pical of the typhoon season is probable. Captain 

 Ault recorded these meteorological conditions in his 

 progress reports of the cruise (3), and continued with 

 very revealing descriptions of the conditions prevailing 

 for the balance of July. Briefly, on July 3 or 4, the region 

 of cold surface water was entered, with practically one 

 hundred percent of mist, fog, or drizzle thereafter until 

 July 21. The fog or mist was "thick" on most days. On 

 July 14 the winds changed from easterly and southerly to 

 southwesterly, and freshened, driving the ship an average 

 of two hundred miles per day for the next several days 

 in contrast with an average of one hundred miles per day 

 for the preceding interval. Thus, there are three periods 

 having different meteorological conditions, these being 

 June 2 to July 2, July 3 to 13, and July 14 to 21. 



Daily values of the atmospheric-electric elements 

 for the three periods are tabulated in table 3. The sum 

 of the two conductivities, X + + X -, is given for each day, 

 together with the simultaneous value of potential-gradient, 

 G, and the computed value of air-earth current density, 

 i. For each of these sets of data the geographical posi- 

 tion is given and the mean time of observations in Green- 

 wich time. The subscript u is used with the designation 

 for each atmospheric-electric element, to indicate what 

 may be called "unusual" or disturbed conditions. Use 

 will presently be made of this subscript. In the last col- 

 umn of the table are given values of in, the subscript n 

 designating normal or undisturbed conditions, the signifi- 

 cance of which will be brought out in later discussion. 



Examination of table 3 reveals that the data for June 

 2 to July 2 were obtained in the 30° to 40° north latitude 

 belt, those for July 3 to 13 in the 40° to 47° belt, and 

 those for July 14 to 21 between 48° and 53° north. These 

 groups are designated 3a, 3b, and 3c, respectively, for 

 purposes of discussion. For group 3a the mean value of 

 air-earth current density is only 7.4 x 10-7 esu, for 

 group 3b, 9.1 and group 3c, 12.9 x 10-7 esu. When one 

 compares these values with those obtained for April and 

 May, 1929, taking account of the increasingly higher lati- 

 tudes in going from April to July, as in figure 15, one 

 finds that, had fog, mist, or haze not been present in June 

 and July, the value for group 3a might have been expected 

 to be about 11.8 x 10-7 esu, for group 3b, about 12.7 x 

 lO-'' esu, and for group 3c about 13.3 x 10-7 esu. That 

 the curve in figure 15, drawn through the points repre- 

 senting the April to July data, should have the slope shown, 

 appears to be supported by the curve for the September 

 to November data, which has been placed in figure 15 for 

 comparison. The thirty-five sets of observations com- 

 posing the September to November data were obtained 

 at approximately Greenwich midnight, whereas the fifty- 

 two sets of April to July data were obtained between 1.5h 

 and 4.5h, Greenwich time. Evidently the universal diur- 

 nal-variation characteristic and the seasonal variation 

 combined to make the September to November values of 



