DEPARTMENT OF TERRESTRIAL MAGNETISM. 255 



The foregoing method, when applicable, has the advantage of giving con- 

 tinuous curves representing the secular changes for all positions within the 

 area investigated. The results obtained from track intersection, on the other 

 hand, gives only the mean change at a single position over the time elapsed 

 between the two passages, but has the great advantage of yielding a positive 

 result, provided a suitable method is used to eliminate accidental errors, and 

 to make the necessary interpolation to a common point. 



The direct method of obtaining secular variation employed at land sta- 

 tions, namely, that of precise reoccupation of stations at intervals, has not 

 been considered expedient in ocean work because of the delays that would 

 ensue, and because of the necessarily large accidental error to be expected in 

 any single observation. It has been customary, in deriving the preliminary 

 values of annual change in previous discussions, to group observations in the 

 vicinity of the intersection of two tracks, and from a mean suitably derived 

 to interpolate a value of the element for the common point. The number of 

 stations to be used in taking a direct mean is limited by the condition that the 

 mean value must apply to the mean position, which can only be true when the 

 change in the value of the element is a linear function of the distance between 

 the stations; the determination of an acceleration factor by a least-square 

 adjustment using the geographic coordinates is inexpedient because of the 

 usual linear distribution of the stations. To suppl}^ an acceleration factor 

 and thus reduce the probable accidental error at the intersection by using a 

 greater number of observations, the following method was tried. 



A portion of a cruise was selected over which the course approximated a 

 great circle for a considerable distance, so that all stations were generally 

 within a degree of a line computed from their geographic coordinates by use of 

 the slope form of the equation of a straight line. The varying length of the 

 longitude degree was compensated by multiplying that coordinate by the 

 cosine of the latitude. The projected distance, d, of any station from the 

 reference-point on such a line, and its perpendicular distance, p, from the 

 line were readily computed from quantities previously derived. Since the 

 stations were selected so that p was always small, the declinations were 

 adjusted by use of a conditional equation, 



AD = h-\-xd-{-yp-\-zd^-{-wdp 

 involving only five coefficients. 



Applications of this method promise improved values of secular changes at 

 track intersections, and, while the computations are necessarily somewhat 

 more extended than the more approximate methods previously used,^ the 

 improvement compensates the additional labor. 



Latest annual values of the magnetic elements at observatories. J. A. Fleming. Terr. 



Mag., vol. 27, 1.57-160 (December 1922). 

 Data and distribution of magnetic observatories, 1922. J. A. Fleming. Bull. 3, Sect. Terr. 



Mag. and Electr., Internat. Geod. Geophys. Union, 72-74, 1 plate (Oct. 1923). 



These papers give compilations of the most recent annual values of the 

 magnetic elements at observatories distributed over the Earth. The second 

 contains in addition information regarding publications and persons in charge; 

 it has also a chart, prepared by C. C. Ennis, showing the distribution of mag- 

 netic observatories. 



Comments on Weinberg's suggestions for field work. J. A. Fleming. Terr. Mag., voL 

 27, 156, 168 (December 1922). 



This paper is a brief discussion of a procedure proposed ^ by Professor Boris 

 Weinberg in land magnetic-survey work. The field methods followed by the 



1 See Re.s. Dep. Terr. Mag., vol. m, pp. 4.30-433 



' Bori.s Weinberg, On the methodology of finding and representing the distribution of a natural 

 element over a certain region of the Earth's surface, with special reference to terrestrial mag- 

 netism, Terr. Mag., vol. 27, 137-155 (December 1922). 



