Reductions to Standard Instruments 45 



The effect on the horizontal intensity correction to standard, due to possible 

 changes in the distribution coefficients, P and Q, over a period of two years or less, 

 was especially investigated for magnetometer No. 5. This instrument was taken 

 because of its extensive use in field and observatory intercomparisons. The dis- 

 tribution coefHcients resulting from 23 sets of deflections at three distances each 

 made at Washington in January and February 1908, and from a least-square 

 adjustment of the observations at eight observatory stations during February 

 1908 to March 1910 (prior to accident), there being four sets or more of deflections 

 for each of three pairs of distances at the observatory stations, were, P = + 14.59 

 with a probable error of 0.07, Q = 640 with a probable error 6.3. The 

 investigation showed, at least for this instrument, that with proper care in hand- 

 ling the magnets it will suffice for all purposes of field work to adopt mean 

 values of P and Q for the entire period. It is of interest to note, however, that in 

 spite of all possible care bestowed in the construction of the magnets of this mag- 

 netometer according to theoretical requirements as to dimensions, etc., the distri- 

 bution coefficients were found considerably different from the theoretical values; 

 in this connection see also tabular details for magnetometers of types 1(c) and 1(&). 



Dip-Circle Corrections to Standard. 



In the regular dip or inclination observations, the polarity of needles is invariably 

 reversed and hence the so-called balance error due to eccentric position of center 

 of gravity of the needle is eliminated. There remains, however, the error due to 

 irregularity of figure of pivot and this will vary, in general, with the magnetic field. 

 Hence the determinations of needle corrections at a base station, however carefully 

 executed, may not necessarily apply to a region far remote where the dip and 

 intensity are considerably different from those at the base station. Use has there- 

 fore been made not only of the comparisons at Washington, but also of all available 

 field and observatory comparisons and a tentative relation of the following form was 

 established whenever warranted, F being the total intensity, I, the inclination, and 

 X, y, z, three coefficients to be determined: 



F^.1 = .r -f- s cos I + y sin / 



An experimental attempt was made at Washington to determine the coeffi- 

 cients, X, y, and s, for the required range by making use of artificially produced 

 magnetic fields.' This method, however, has been abandoned at present because 

 of the great expenditure of time required and because experience has shown it to 

 be better to utilize results obtained under actual conditions of field work and 

 extending over the entire period involved. In very exceptional instruments it 

 was possible to secure some data for corrections by studying critically the differ- 

 ences exhibited by needles among themselves for the range of dip encountered. 



The dip corrections adopted in this volume are given separately for each instru- 

 ment; they are to be applied algebraically regarding dip, north end down, as positive 

 and south end of needle down, as negative. For type of instrument referred to, 

 see section on "Instrumental Equipment of Parties." 



CJ. Dike, P. H. Experimental investigation of dip needle corrections. Terr. Mag., v. 14, 1909. (137-146.) 



