248 



Mr. C. Chambers. 



Kew Observatory.* The records extend, in the present case, from 

 June 1, 1847, to December 31, 1872, and consist of differences (always 

 taken to be of positive sign) between the highest and the lowest 

 values of easterly declination observed by Grubb's declination mag- 

 netometerf on every observation day of the period. Until the end of 

 the year 1865 the observation day was the Gottingen astronomical 

 day ; after that time it was the Bombay civil day. The daily differ- 

 ences were obtained from hourly observations made on all days 

 except Sundays and a few holidays in each year. Grubb's declination 

 magnetometer is of the well known form described in the Report of 

 the Committee of Physics of the Royal Society, 1840 (p. 13). Up to 

 1868*00 each individual entry in the register of the scale-reading of 

 the instrument was at once converted into easterly declination in 

 minutes, and the daily ranges are the differences of such converted 

 readings, but after the date named the differences of the scale- 

 readings were first taken, and then converted into minutes. The 

 ranges include the effect of disturbance. 



A. Annual Variation of Declination-Uange. 



2. The year being divided into forty-eight equal parts, commencing 

 with the midnight between the 31st December and 1st January, J means 

 were taken of the ranges for the fifteen days preceding and fifteen 

 days following the nearest midnight to the commencement of each 

 48th part of the year. Attributing four of the 48ths of a year to each 

 month, and designating as " monthly means " the thirty-day means 

 thus obtained, the following table exhibits each of the forty-eight 

 results for each year, and on the average of all the years : — (See pp. 

 250 and 251). 



The numbers in the last column are taken to represent the annual 

 variation — combined with the annual mean value — of declination- 

 range. 



E. Variation of Long Period. 



3. Proceeding now on Dr. Stewart's hypothesis as to the relation 

 between declination-range and solar activity, we divide the numbers 

 in each line of Table I by the mean number (in the last column) of 

 that line, and multiply the quotient by 1000, thus obtaining a table 

 " exhibiting the monthly means of declination-range (forty-eight 

 points to each year), the mean value of the range for the whole series 



* "Proc. Eoy. Soc," vol. 26, p. 102. 



f On the rare occasions when this instrument was under adjustment, the blanks 

 in its register were filled up from the register of a small declination magnetometer 

 which was used as a subsidiary instrument. 



X Leap-years were taken to contain 366 days, and other years 365 days. 



