184 



DISCUSSION OF THE MAGNETIC OBSERVATIONS. 



BY 



DR. C. CHREE, F.R.S. 



1. The observations taken by Mr. MOSSMAN in Laurie Island, South Orkneys, were reduced by the 

 staff of the Observatory Department of the National Physical Laboratory. They consisted of absolute 

 observations of declination, inclination, and horizontal force, made between May, 1903, and February, 

 1904, and of hourly readings of declination during ten days in December, 1903, and ten days in 

 January, 1904. 



2. Declination. The observations were taken with a Unifilar Magnetometer No. 2, by the Cambridge 

 Instrument Company, the absolute observations being always made with the scale both erect and 

 inverted. The readings were referred to a distant mark, described in Mr. MOSSMAN'S notes, whose 

 azimuth relative to the geographical meridian was determined by a number of sun transit observations. 

 The sun observations were reduced at the Argentine Magnetic Observatory at Pilar, the mean value 

 obtained for the azimuth of the mark from ten observations being 9 15' west of north. The conditions 

 under which the sun observations were taken were not very favourable, and individual determinations of 

 the azimuth of the mark differed by a few minutes of arc ; the mean result, however, should not possess a 

 large probable error. 



Table I. gives the date and the mean (local) time of the individual observations, as well as the resulting 

 values of the declination. The observations on June 15, and July 2, 3 and 12, appear abnormal, and have 

 been omitted in forming the monthly means. No attempt has been made to allow for the diurnal 

 variation. The observations were made, with few exceptions, between 11 a.m. and 7 p.m., and, as we 

 shall see later, easterly declination was above the mean during this portion of the day, at least near mid- 

 summer. Thus the correction to the mean value for the 24 hours would almost certainly cause a reduction 

 of the figures in Table I. in the great majority of cases. The diurnal range, however, as we shall see 

 later, is not large even at midsummer, when it is usually about its maximum, and the corrections that 

 remain to be applied to the monthly means in Table I. would probably, in most cases, not exceed 2 or 3 

 minutes of arc. The absolute observations themselves would suffice to show that the range of declination 

 is not large, for, excluding the four observations already mentioned as doubtful, we have for the largest 

 and smallest of 40 observed declinations 5 35' -6 and 5 24'- 4 respectively, the difference being 

 only 11' -2. 



If we assign to May 31 the mean 5 33' 6 of the monthly means for May and June, and to January 31 

 the mean 5 29' - of the monthly means for January and February, we obtain a decrease of 4' -6 in eight 

 months, i.e., of 6' - 9 in the year. There may, however, be a considerable annual period (i.e., a regular 

 change whose period is 1 year), and there are other sources of uncertainty, so that much weight cannot be 

 assigned to this estimate of the rate of secular change. 



3. Inclination. The observations of inclination were made with the Barrow Circle No. 24, having two 

 needles, Nos. 1 and 2. On most occasions observations were made with both needles. The results from 

 the two needles did not differ much, but on the average the inclination obtained with No. 2 exceeded that 

 obtained with No. 1 by 0' 53. On the few occasions when observations were made with No. 1 only, an 

 imaginary mean for the two needles has been obtained by adding 0' 3 to the result obtained. 



T;il>lc II. gives full particulars of the individual observations. The monthly means and the mean for 

 the whole series are based on the mean results for the two needles. The mean values obtained for July 

 and August are lower than those for either the earlier or the later months, and no deduction seems 

 possible as to either the secular change or the diurnal variation. 



The difference 9' 7 between the extreme values 54 35' 3 and 54 25' 6 is, relatively considered, very 

 considerably larger than the corresponding difference in the case of the declination. 



4. Horizontal Force,. Table III. gives particulars of the observations of horizontal force H. On some 

 occasions the interval elapsing between the vibration and deflection experiments was a little long. In 



