254 CARNEGIE INSTITUTION OF WASHINGTON. 



this tendency is real. In the case of the declination, the evidence is positive. 

 The two stations which were occupied in August 1907, in January 1910, and 

 in August 1922 were found to have declinations which differed by 3° 53', 

 3° 59', and 4° 18' on the three occasions respectively, showing a steady increase 

 in the difference with time. This increase seems to be related in some way to 

 the declination at the two stations and is suggested also by the results of the 

 other stations; for if the 15-year change is plotted for each of the five stations, 

 using the declination as abscissae, the points fall near a line within limits on 

 the order of the uncertainties of determination. It is hoped that a further 

 refinement of intensity observations may make it possible to reveal a connec- 

 tion between this phenomenon and the components of the disturbance field. 



Note on the determination of magnetic secular variation from ocean observations. H. W. 



risk. 



Whenever it is permissible to assume (a) that a magnetic element varies 

 uniformly or with a uniform acceleration along each meridian and along each 

 parallel, (b) that the annual rate of change of the element at any point is 

 constant or has a uniform acceleration, and (c) that the annual rate of change 

 varies uniformly from point to point along each meridian and along each 

 parallel, then the geographical distribution and the secular variation of that 

 element may be expressed by an empirical formula of ten coefficients, con- 

 sisting of a constant term, the first powers, squares, and second-power products 

 of the differences between the position and epoch selected for reference and 

 the latitude, longitude, and date considered. 



While these three fundamental conditions apply strictly only to very small 

 areas on the Earth's surface, they may be assumed to hold approximately 

 for one or fnore of the magnetic elements over comparatively large areas in 

 certain regions. 



The section of the Atlantic Ocean lying immediately east of the United 

 States and west of the meridian 45° west longitude has been traversed by the 

 Carnegie on each outward voyage and return to home ports, so that observa- 

 tions are well distributed over this area, and have been made at various 

 times between 1909 and 1921, thus presenting favorable conditions for testing 

 results of adjustments by the foregoing formula. Such adjustments were 

 made for all three magnetic elements in connection with the reduction and 

 investigation of the Bermuda magnetic anomaly (see pp. 253-254). 



The stations at which horizontal intensity and inclination were observed, 87 

 in number, were combined into 39 groups, so as to preserve as far as possible 

 the maximum difference in time in each portion of the region covered and to 

 make the groups of approximately the same weight. The resulting equations 

 for declination and inclination give changes for two stations on the Bermuda 

 Islands from 1910, when they were occupied by the Carnegie party, to 1922, 

 when they were reoccupied by a land party, of —9327 in intensity and +30' in 

 inclination. The corresponding changes determined independently from the 

 land observation are —9447 and +33', an agreement well within the Umits 

 of determination. 



The adjustment of the declination observations in the same area required 

 a different grouping, as they were made at different times of day and were 

 more numerous than those for intensity and inclination. In all, 129 stations 

 were combined into 40 groups. Later, those groups lying north of parallel 32° 

 N. and west of meridian 56° W. were adjusted separately, and finally an 

 adjustment was made of the groups within one degree of parallel 38°, assum- 

 ing the change along the meridians for the short distance involved to be that 

 obtained from the earlier discussions. The residuals resulting from these 

 adjustments were not satisfactory, the difficulty arising from the fact that 

 the third of the three fundamental conditions is not fulfilled. 



