Reduction Formula and Instrumental Constants 



237 



Sea deflector 3 (D3). — Sea deflector 3 of the revolving-compass pattern, designed and 

 constructed by the Department of Terrestrial Magnetism, is described on pages 191 and 192 

 and illustrated on Plate 12, Figure 1. It was used on Cruises I and II (as far as Cape Town, 

 March 1911). Since the courses followed during this period were such that the vertical 

 intensity practically varied uniformly with the tiaie, between successive shore determina- 

 tions of mC, a graphical adjustment of the available data, referred to the standard tem- 

 perature, 20° centigrade, was found to give, with sufficient accuracy, the value of this 

 constant for each magnet and distance at any time. The temperature factors were deter- 

 mined as explained above. An examination of the data indicated that there were no 

 periodic corrections to log mC, as was the case for the less-accurately made deflectors 1 

 and 2 used in the magnetic work on the Galilee. 



The constants adopted on the basis of C. I. W. Standard (see p. 232) are given by the 

 following equations, which ai'e to be used in connection with the values of log mC at 20° 

 centigrade, adopted from the graphical adjustments, and given in Table 56 for different 

 dates: 



Magnet 45 log mC = log mC at 20° for t -hO . 00026 (20° - 1) 

 Magnet 2L log mC = log mC at 20° for t+ 0.00014 (20°- 



Table 56. — Logarithms of Intensity Constants at SO" Centigrade of Sea Defkctor 3. Cruises I and II (to March 1911). 



'The deflection distances 1 and 3 only were used for observations at sea. 



Sea deflector 4- — Sea deflector 4 of the revolving-compass pattern, with numerous 

 improvements on deflector 3 in mechanical detail, designed and constructed by the Depart- 

 ment of Terrestrial Magnetism, is described on pages 192 and 193 and illustrated by Figures 

 2-9, Plate 12. It was used on Cruises II (from April 1911), III, and IV; during April and 

 May 1914, preceding Cruise III, it was thoroughly overhauled and repaired. A sUght leak 

 developed in the inner lining of the bowl during Cruise II, and again during Cruise IV, 

 but did not affect the intensity constants. It appears that some change, of unknown 

 cause, took place in magnet 45 just before the comparison observations at Antipolo in 

 February 1912; that the change occurred at Antipolo is borne out by comparisons of the 

 sea values of H before and after this station, obtained separately from observations with 

 the two magnets 45 and 2L. 



The adopted constants for Cruise II from April 1911, on the basis of C. I. W. Standard 

 (see p. 232), resulting from least-square adjustments of all the available data, are given in 



