276 Reports on Special Researches 



more than a ie\x tenths of a niinute; in some instances M-here the conditions were not good, 

 when, for example, it was not possible to exchange stations because it was not feasible to 

 observe with the comparing instrument at precisely the same station where the compared 

 instrument was mounted, the error in aD or Al may be as much as 0'.5 or more depending 

 upon the precise circumstances. There is no question, however, that the declination or 

 inclination standards at some observatories may be in error by as much as 1'. It is unfor- 

 tunate that rather too frequently those who are intercomparing observatory standards fail 

 to get declination results. Our experience has shown that it can not be assumed that the 

 instrumental declination-differences are neghgible quantities, if one strives to secure the 

 attainable accuracy in magnetic work. It is also of interest to know that to obtain the same 

 degree of accuracy in the determination of the vertical intensity, Z, as appears possible with 

 the very best magnetometers iorH about 0.00015 to 0.00010 part it is necessary to deter- 

 mine the inclination, /, at observatories in medium magnetic latitude, to about 0'.2. 



But few can realize the very large amount of work and patience involved in securing 

 the comparison data set forth in Table A. In no instance have the comparisons been 

 obtained as the result of special expeditions, but invariably in the course of an observer's 

 field work, and with the instruments used by him in the field. A critical examination of 

 the results will show that they compare favorably with those obtained by others, whose 

 special mission it was to intercompare observator}- standards. There is thus afforded 

 evidence not onh' that the observers of the Department of Terrestrial Magnetism performed 

 their work faithfully and zealously, but also that the observatory authorities did their 

 utmost to perform their part. It is a great pleasure to acknowledge here our obligations 

 to the directors and observers at the various observatories for the courtesies shown and 

 the assistance rendei'ed. 



Table B (p. 278) contains the preliminary results of indirect comparisons between the 

 International Magnetic Standards and 8 magnetic observatories not contained in Table 

 A, derived as explained on p. 270 (Table 31 C.) 



Table C (p. 278) is an extract from Table A. 



An approximate attempt was made to indicate in the columns, "Grade of Value," of 

 Tables A and B, by the letters a, b, c, and d, the degree of reliabiUty of the quantities aD 

 (declination difference), aH H (horizontal-intensity difference), and A/ (inchnation differ- 

 ence) ; a stands for the highest grade of reliability, and d for the lowest. In determining 

 the grade, as full consideration as possible was paid to the conditions under which the 

 results were obtained; whether there were any independent checks; whether station- 

 differences were eliminated, etc., etc. The grade c was assigned to all the results in Tabl<? B, 

 as they depend upon the preliminary reports of Messrs. Dubinsky and Kiihl and are, accord- 

 ingly, subject to future correction. 



In the three columns, under the general heading of "I. M. S. Observatory," are given 

 the resulting algebraic differences for the three magnetic elements, declination (D). hori- 

 zontal intensity (H), and inclination (/), between the adopted international magnetic 

 standards (I. M. S.) and the observatories named in the second column, the standards of 

 which are designated in the last two columns. It must be remembered that the quantities 

 apply only for the instruvicnts and constants as used at the time of the comparisons. 



WHAT ARE THE INTERNATIONAL MAGNETIC STANDARDS? 

 It has been explained in this report how, by more extensive and more varied inter- 

 comparisons of magnetic instruments of all kinds than have ever been carried out before, 

 we have arrived at what have been termed "International Magnetic Standards." It has 

 been shown that during the period of our preliminary investigations these standards may 

 be regarded as having remained constant within all practical requirements. The provi- 

 sional international standards for decUnation and horizontal intensity, chosen in 1907, were 

 found to agree with the international standards of 1914 within 0'.2 in declination and 



