MAGNETIC OBSERVATIONS 291 



March 20 to April 16, 1904 : 



D = 22° 43/7 + 27/93 sin {e + 0° 26') + 7.'96 sin (21? + 274° 24') + 

 2/67 sin (3 e + 149° 03') + 3/56 sin (4 <> + 65° 34') ± 0/39 



April 17 to May 14, 1904 : 



Z? = 22° 37/1 + 25/51 sin (<!'+353° 22') + 8/81 sin (2^ + 287° 10') + 

 3/37 sin (31? + 224° 52') +3-'8i sin (4^ + 24° 14') ±0/27 



May 15 to June 11, 1904: 



D= 22° 40/1 +3o/3i_ sin (<? + 35o'' 15') + 9-'H sin (2^ + 308° 51') + 

 6/64 sin (3(5' + 223° 42') + 1/50 sin (4^ + 31° 56') ± 0/39 



June 12 to July i, 1904 : 



Z>=22° 32/1 +41/83 sin (<? + 34r° 19') + 8/65 sin (2 (?+ 309° 04') + 

 4-'57 sin (3 <> + 1 56° 34') + 3/88 sin (4 e + 100° 30') ±0/53 



For whole observational series, October 4, 1903, to July i, 1904 : 



D = 22° 42/0 + 20/52 sin {B + 357° 14') + 7/69 sin (2 ^ + 294° 16') + 

 1/50 sin (3e+i77" 59') + i-'5i sin (4(? + 47" 52')±o/i2 



For the year, July i, 1903, to June 30, 1904, from monthly grouping of observations and 

 interpolations therefrom : 



Z>=22° 38/5 + 21/11 sin (^ + 354° i8')+6/90sin (25 + 291'' 26') + 

 1/75 sin (3«+ 156° 48') + 1/62 sin (46' + 58° 51') ±0/20 



Month of July, 1903 (interpolated values) : 



Z>=:22° 24/9 + 32/62 sin (5+341° 59')+5.'95sin (25+291° 36') + 

 4/79 sin (35+ 154° 30') + 2/73 sin (45 + 84° 45') ± 0/75 



Month of August, 1903 (interpolated values) : 



D = 22° 25/8 + 24/66 sin (5 + 343" 19') + 4/26 sin (2 5+ 273° 14') + 

 3/59 sin (35+116° 51') + 2/63 sin (45 + 86° 040+0/90 



Month of September, 1903 (interpolated values) : 



D = 22" 30/9 + 17/95 sin (5 + 351 ° 52') + 5/09 sin (2 5+ 272° 49') + 

 2/25 sin (35 + 72° 56') + 2/28 sin (45 + 90° 30') ± 0/66 



In all of these expressions the angle 5 counts from 15° as 0.5 hour a. m. local mean time. 

 The formulae have been carried to terms of the fourth order only, as little or no improvement 

 results from the further extension of the function. Graphical representations to scale of these 

 formulEe, together with the quantities from which they are deduced, are given by figures 5 to 19. 

 , A comparison of the diurnal variations — represented by the four sine terms of the analyt- 

 ical expressions — ^with the observed quantities at the mean local half hours, is given by the 

 following tabulation wherein the observed, interpolated, and computed values on mean of day 

 are indicated by the letters O, /, and C respectively. As a criterion of the accuracy of the 

 formulae for diurnal variation, a column showing difEerences of observed and computed varia- 

 tions is added (XO-C) or {/-€)). In these tabulations a plus sign indicates the variation to 

 be to the east of mean declination for day, while a minus sign indicates the variation to be to 

 the west of mean declination for day. The extreme values on the half hours for each period, 

 both observed and computed, are indicated by bold-face type. 



