70 CARNEGIE INSTITUTION OP WASHINGTON. 



In spite of the apparently vast accumulation of data, such an 

 important question as whether the earth's magnetic energy is 

 increasing or decreasing and the annual rate of change can not be 

 definitely answered. The chief reason for this unfortunate state of 

 affairs is that the accumulated material has not the required general 

 distribution, but pertains chiefly to civilized and restricted land 

 areas, leaving almost neglected the greater part of the earth covered 

 by water. Systematic magnetic survej^s of the oceanic areas are 

 entirely lacking, such results as are at hand having been obtained 

 from occasional expeditions or incidentally to other work. There is 

 here revealed to the Institution a most useful and promising field of 

 work, and definite recommendations pertaining to this matter will 

 be given in a separate communication. 



The completion of the critical study of the modern magnetic charts 

 furnished the necessary data for drawing the following conclusion of 

 great interest in terrestrial magnetism, atmospheric electricity, and 

 meteorology, viz : 



All of the modern magnetic charts — /. e., since those of 

 Sabine for 1840-45 — unite in indicating the probable existence 

 of vertical earth-air electric currents of the average intensity 

 over the region 45° N. to 45° S. of ^ of an ampere per square 

 kilometer of surface. These currents of positive electricity pro- 

 ceed upward (from the earth into the air) near the equatorial 

 regions, where there are ascending air currents, and downward 

 near the parallels 25° to 30°—/. r. , in the regions of descending 

 air currents. Near the parallels 40° the electric currents are 

 again upward, thus corresponding once more with the general 

 atmospheric circulation. Beyond the parallels 45° the results 

 appear too uncertain to warrant drawing a definite conclusion. 



In order to make some tests as to the manner of distribution of 

 the upward and downward electric currents, the currents over quad- 

 rilaterals bounded by two parallels 10° apart and two meridians, 

 likewise 10° apart, have been derived for the entire region from 

 60° N. to 60° S. for the three epochs 1842, 1880, and 1885. As a 

 general result it did not appear as though the directions of the 

 electric currents — whether up or down — were to be associated with 

 the distribution of land and water. There was, however, a decided 

 indication, Air each epoch, that over the areas of low pressure, where 

 the air currents are upward, there the electric currents were likewise, 

 in general, upward, and that over the areas of high pressure, where 

 there are descending air currents, there the electric currents were 

 likewise descending. 



