Comparisons of Magnetic Standards, 1915-21 



409 



edly for heavy current for other laboratory use. Each value on June 3 and 4 by mag- 

 netometer No. 3 (correction on I. M. S. being zero for constants of December 12, 1910, 

 see p. 10), depends upon 2 sets of deflection observations only at 3 distances using a 

 mean value of magnetic moment determined from a number of complete observations 

 both preceding and following the comparisons. For the work in August each I. M. S. 

 value depends upon a complete determination by magnetometer No. 3, i.e., 2 sets of 

 deflection observations at 3 distances preceded and followed by oscillation observations. 



Table 20F. Results of Horizonlal-Intensily Comparisons behveen C. I. W. Standard Magnetometer No. 3 and C. I. W. 



Sine Galvanometer No. 1, at Washington, 1921. 



Series 



Date 



1921 

 June 



Local mean time 



From 



To 



Hor. int. obtained 1 



I. M. S. 



y 

 1S720.4 



695.6 



661.8 

 669.9 



S. G. 1 



Weighted mean value of (I. M. S. S. G 1). 



II< Aug. 2 

 2 



4 

 4 

 4 

 4 

 6 

 5 

 5 

 5 

 8 

 8 

 8 

 8 



12 03 

 14 55 



9 43 

 11 42 



13 29 



14 30 

 9 42 



11 39 



13 28 



14 25 



05 



1 50 



3 03 



4 51 



14 37 



15 50 



11 18 



12 26 

 14 07 



16 01 



11 14 



12 22 

 14 09 

 16 16 



1 40 



2 32 



4 38 



5 40 



18689 

 711 

 663 



679.6 



686.6 

 638 



660 



669 

 675 

 672 

 678 

 665 



18720.2 



696.0 



662.2 

 669.5 



1S693.6 

 708.0 

 659.6 



676.4 



690. S 

 638.4 



661.5 



667.4 

 674.9 

 679.3 

 680.6 

 673.1 



Weighted mean value of (I M. S. S. G. 1) 



Weighted mean value 5 of (I. M. S. S. G. 1) from I and II . 



I.M.S.- 

 S. G. 1 



+0.2 



-0.4 



-0.4 

 +0.4> 



Weight 



0.5 



0.5 



0.5 

 0.2 



Station 



M. 3 S. G. 1 



iV, 



-O.I7 or -0.00001// 



iV 



AT 



-O.87 or -0.00004// 



-0.7-ror -0.00004// 



' All values are referred to station N m using station-difference determined from each series, viz: for I, Nm = Sn-h6.6yS 

 for II, N m = S+5Qy. 



1 Each value given in series I for magnetometer No. 3, except as noted, is from 2 sets of deflections only at 3 distances 

 using value of magnetic moment interpolated from extended complete determinations preceding and following this work. 



Half set. 



Each value given in series II for magnetometer No. 3, except as noted, is from a complete determination, viz, oscil- 

 lations, deflections at 3 distances, second set of deflections at 3 distances, and second set of oscillations. 



Probable error of the weighted mean value is 0.67 or 0.00003//, and that of a single determination of weight one 

 is 2.07 or 0.00011//. 



The conditions for comparison work were by no means good as moderate magnetic 

 storms were in progress on June 3, p. m., to June 4, on the afternoons of August 2 and 3, 

 on August 5, and during the early morning of August 8. As has already been stated 

 there are slight effects occasioned by the electric car-lines about % mile west of the 

 Standardizing Magnetic Observatory; these apparently are eliminated in the mean of 

 observations taken over the short period required for a set of oscillations or of deflec- 

 tions by the magnetometer or for a complete determination by the sine galvanometer. 

 The agreement of I. M. S. and C. I. W. sine galvanometer is surprisingly close and indi- 

 cates that the standard in horizontal intensity adopted for the discussion of various 

 results in Volume II, pages 270 to 278, is correct within one part in 10,000." The prob- 



It is merely a coincidence, but nevertheless interesting, that in a special report by Messrs. John A. Fleming and 

 Harlan W. Fisk, not yet published ("Discussion of magnetometer-corrections on standards in declination and horizontal 

 intensity") investigating, among others, the long series of observations during 1907 to 1921 by C. I. W. standard magneto- 

 meter No. 3, the final values for mean distribution coefficients, mean inertia, and correction because of bending of deflection 

 bar, indicate that values by No. 3 should be increased by 0.00004//. 



