274 Reports on Special Researches 



Let us cite a few cases. As far as can be determined, from numerous intercomparisons, 

 the relations between the D and H standards of the United States Coast and Geodetic Survey 

 and those of the Carnegie Institution of Washington (Department of Terrestrial Mag- 

 netism) have remained unchanged, during the period 1905-14, within about O'.l or 0'.2 in 

 declination and within about O.OOOli/ (27) in horizontal intensity. The same degree of 

 accuracy, however, can not be said to appl}^ to the /-standards of the two respective organi- 

 zations. As will be seen from Table 7C, p. 229, the difference (C. I. W. Cheltenham) 

 has apparently changed by about 1' between 1910 and 1913; whether this is to be ascribed 

 to some extraneous source of disturbance in the later observations, or whether it indicates 

 some change in the large stationary Wild-Edelmann earth inductor has not yet been 

 definitely settled. From the many comparisons at Washington between the C. I. W. 

 standard and numerous dip circles and earth inductors, no change in the C. I. W. standard 

 greater than O'.l or 0'.2 during the period 1907-14 is indicated. Furthermore, up to 1910 

 there was no indication of any change in the relation of the C. & G. S. and the C. I. W. 

 standards for / amounting to more than O'.l or 0'.2; the cause of the apparent difference 

 of 1' in 1913 is under investigation. (See footnote, p. 226.) 



The various checks applied (see pp. 242-243, 252, 253) for strengthening the determi- 

 nation of relations between the C. I. W. standards for D, H, and I, and those of Kew and 

 Potsdam two observatories where, it is known, every care is bestowed upon instruments 

 and constants show no indications of changes in the respective standards greater than can 

 he accounted for by purely observational error. Further evidence of the apparent constancy 

 of magnetic standards is obtained from the repeat comparisons at observatories where no 

 change has been made in the observatory instruments and constants. (See, for example, 

 Agincourt, Table 1 C, p. 215; Helwan, Table 12G, series II and III for H and 7, p. 238; 

 Rome, Tables 22A, 22B, 22C, series I and II, pp. 255-250.) It should be remarked that 

 in these repeat comparisons, severer tests have been applied than need be the case for 

 purely inter-observatory comparisons; that is, the intermediary C. I. W. instruments with 

 which the comparisons between the observatory and the C. I. W. standards were effected, 

 have generally been different in the two series of comparisons and, in fact, have been stand- 

 ardized at Washington at different times and have been subjected to severe conditions of 

 field work. 



It will be interesting to give, from our accumulated data, some examples showing the 

 apparent constancy of field instruments in spite of exceptionally severe conditions of field 

 travel. On p. 269, Table 30 D, an exhibit has already been made of the constancy of C. I. W. 

 magnetometer No. 5 for the period 1908-10, during which difficult magnetic-survey work in 

 Turkey, Asia Minor, Persia, and Egypt was accomplished, and intercomparison data were 

 secured at 5 magnetic observatories. (See the close agreement in the resulting corrections 

 on the C. I. W. standard as determined by direct comparisons at Washington in December 

 1907 and February 1908, and by indirect comparisons in the field in January 1910; see also 

 the close agreement in the results of the Kew repeat comparisons, Tables 14 A, 14 B, series 

 I and II, p. 241.) 



C. I. W. magnetometer No. 7, used in extensive and strenuous field work in 1908-13 

 in North America, Central America, Europe, Asia, and Africa, affords another instructive 

 example; in some of this field work it was necessary to travel in a small sailing boat, and 

 often landings had to be made through breakers or surf. 



Considering the extent and character of travel in the intervals between the various 

 comparisons at Washington, the small changes in the quantities aD anil \H 'H, shown in 

 Table 31 F must l)e regarded as highly satisfactory. 



