Magnetic Observations ox the Carnegie 259 



The application of variation corrections to the observed results on account of the 

 numerous variations of the Earth's magnetism, e. g., diurnal variation, secular variation, 

 magnetic perturbations, etc., is deferred to the volume in which all the magnetic data, 

 obtained both on land and sea, are summarized and reduced to a common epoch. (That 

 volume, probably No. V, can not appear until some time after the completion in 1917 of the 

 Carnegie's present cruise. Whether it will be worth while, in the case of the ocean data, to 

 apply any other corrections than those on account of secular change will there receive 

 consideration.) To avoid undue delay in the promulgation of the accumulated data, and 

 in view of the inaccuracies of the magnetic charts at present in use, it is considered best to 

 publish the observed results as obtained with no corrections appUed except the reductions 

 to magnetic standards, as fully explained in the section on this subject (see pp. 232-256). 

 However, since for the magnetic elements tabulated the precise date and local mean time 

 of each observation are given, the reader is suppUed with the requhed information in case, 

 for some purpose of his own, he desires to reduce the observed values to some mean time. 



Combining Weights Assigned to Different Instruments and Methods. 



The tabulated values of the magnetic elements are the weighted means, usually of 

 two or more results, obtained with two different instruments, or by two different methods. 



To obtain the weighted mean value of the decUnation, the results with the standard 

 compass (marine coUima ting-compass, CI) were given a combinmg weight 2, whereas the 

 auxiUary results with sea deflector (D3, D4) received the weight 1. 



The weighted mean value of the inclination was obtained by assigning the weight 2 

 to the result from each dip needle and the weight 1 to the result derived from each complete 

 observation of deflected dip. Hence, the inclination results from long and short distance 

 each received a weight of 1, or if the observation at one distance was repeated, the result 

 was given a weight of 2. At the stations where the incUnation was determined both with 

 the dip circle and the earth inductor, the dip-circle result, obtained as just described, 

 was, in general, combined with the earth-inductor result by giving equal weights to the 

 two instruments. When these two results differed by more than 0?2, the dip circle was 

 given weight 2 and the earth inductor weight 1. WTiile the earth inductor on land gives 

 results superior to those of the dip cu-cle, certain difficulties enter in marine-inductor 

 work which have not yet been entirely overcome. 



The weighted mean value of the horizontal-intensity results was obtained by assign- 

 mg weights 3, 2, and 1 to the sea-deflector results, the sea dip-circle results by deflections, 

 and the sea dip-cu-cle results by loaded needle, respectively, when the various results were 

 obtained under normal sea conditions. But when the observations were made under 

 unfavorable conditions of motion or with small values of horizontal intensity, the weights 

 assigned were then 6, 4, 1 , in the order designated. In some exceptional cases equal weights 

 were assigned the results obtained by sea deflector and by sea dip-chcle (deflected dip or 

 loaded dip), as m the case of swings, exceptionally quiet conditions, etc. 



The weights referred to above are not to be confused with the figures which appear in 

 the "Wt." columns of the Table of Results. The tabular weights refer to the conditions 

 as to sea and weather under which the observations were made (see p. 258). 



EXPLANATORY REMARKS FOR PRELIMINARY RESULTS. 1915-1916. CRUISE IV. 



To meet the requests received from various hydrographic establishments, it has been 

 decided to give in this volume, in addition to the final results of the ocean magnetic work 

 on the Galilee and the Carnegie, 1905-1914, the preliminary results for the subsequent 

 work. These preliminary values of the magnetic elements are derived from computations 

 made and checked aboard the vessel, and are dependent upon preliminary values of instru- 

 mental constants. Accordingly, they are subject to future revision when the ofl^ce com- 

 putations are made with the final instrumental constants. It is not probable, however, 



