416 Reports on Special Researches 



Now, in so far as the conductivity is proportional to the ionic density, and the latter 

 is proportional to the square root of the rate of production of pairs of ions, we should expect 

 that if q is the value of the latter quantity for the Pacific Ocean, and q+dq the value for the 

 Atlantic Ocean, 



(1±^)*=1.07 



or approximately dq'q = 0.14. 



Taking the value 5 = 1.6 found on page 4 14, for the great oceans, as the value for the Pacific 

 Ocean, we thus find 5^ = 0.22. As already stated on page 413, the mean value of the activity, 

 expressed in Elster-and-Geitel units, for several land stations is 85, so that if this be taken 

 as corresponding to the mean emanation content, 88X10"^^ curie per cubic meter, for the 

 land,' we find that 1 Elster-and-Geitel unit corresponds to l.OX 10"''^ curie of emanation per 

 cubic meter, and consequently, according to page 414, to a rate of production of 0.021 ion 

 per c. c. per second.^ In order to account for the above value 0.22 for dq, it would conse- 

 quently be necessary to assume, over the Atlantic Ocean, a radioactive content which was 

 in excess of that over the Pacific Ocean by about 10 Elster-and-Geitel units. Taking the 

 mean value for the Pacific Ocean as 3 Elster-and-Geitel units,' we should, on these lines, 

 expect 13 Elster-and-Geitel units over the Atlantic Ocean, and this is just about the order 

 of magnitude of the radium-emanation content found there. Thus Hewlett found 12 and 

 Johnston 23 Elster-and-Geitel units respectively, in the north Atlantic Ocean, on the 

 Carnegie's second and third cruises; the mean of these values is 18. 



It is noticeable that the slight difference between the average values recorded in Table 85 

 for the corresponding ionic contents and conductivities in the Pacific and sub-Antarctic 

 oceans is in the right direction to be accounted for by the difference in the emanation 

 content. The mean of the values of n+ and ?i_, recorded in Table 85 for the Pacific Ocean, is 

 752, and the corresponding value for the sub-Antartic oceans is 722. If q is again taken 

 for the rate of production of pairs of ions per c.c. in the Pacific Ocean, and ii q—8q is the 

 corresponding value for the suId- Antarctic oceans, we readily find, on the hne of the argument 

 given above, 



dq = O.OSq 



so that taking g = 1.6 as before, 8q = 0.13. This corresponds in radium-emanation content to 



(0.13/0.021)X10~'^ = 6X10~'^ curie per cubic meter. The difference between the Pacific 



and sub- Antarctic emanation contents amounts, according to Table 85, to 2.9X10"'^ curie 



per cubic meter, and is thus of the right order of magnitude to account for the differences 



in the corresponding ionic contents. An exact numerical agreement can not, of course, be 



expected. 



DIURNAL VARIATION. 



As already stated, diurnal-variation measurements were made for the quantities 

 X, 11+, and R, but it was naturally not practicable to make complete diurnal-variation 

 observations very often. In so far as a diurnal-variation curve may only be expected to 

 approximate to a definite and characteristic form when the results of many sets of observa- 

 tions are combined, it was considered best to combine, into one curve, for any element, all 

 of the corresponding diurnal-variation observations throughout the cruise. A curve ob- 

 tained in this way consequently corresponds to the mean diurnal-variation curve throughout 

 the period of the cruise. Although the observations here discussed extended over about 



'See page 413. 



'Kurz has made a direct comparison of the Elster-and-Geitel unit with the corresponding amount of ionization to be 

 expected. His data have, at the hands of Kohlrausch and others, suffered various corrections which have changed the 

 original result by 260 per cent of its value, and the corrected value gives the rate of production of pairs of ions per c.c, 

 corresponding to 1 Elster-and-Geitel unit as 0.029. (See K. Kurz, Abk. Ak. Wiss. Maih.-phys. Kl.. No. 1, vol. 25, p. 44, 

 Munich. 1909; also K. W. F. Kohlrausch, Phys. Zeit., vol. 13. p. 1193. 1912). 



^See page 413. 



