Values of Declination Base-Line Derived from the means of five consecutive weekly determinations.
January, 22 74'2 July, 9 76'7 29 74'2 16 76·8 February, 5 74'3 August, 6 77°0 12 74'2 14 77°0 19 74·2 26 74'1 March, 5 74'4 September, 3 ...... 12 74·6 10 77'1 19 74·8 17 77·2 26 77·1 24 74·9 April, 2 75'0 October, 1 77'2 9 75'1 9 77·1 16 75'3 16 75·7 23 77·1 22 76·8 May, 1 75'7 November, 5 76·7 7 76·0 12 76·8 14 76·2 19 76·4 21 76·8 26 76·5 28 76·5 December, 1 76·5 June, 4 76·5 11 76·3 15 76·0 17 76·6 22 75·7 25 76·5 29 75·455. A time drift in the horizontal force and vertical force records, which appears to be due to a weakening of the quartz fibres rather than a reduction in the moment of the magnets, has made the determination of the temperature co-efficients a matter of great difficulty.
56. It is expected that a supply of alternating current from the mains of the China Light & Power Co. will be available about the middle of 1932. With this power it is hoped to keep the temperature of the recording room nearly constant. The time drift of both instruments can then be accurately determined.
57. Attempts to increase the sensitivity of the vertical force instrument failed, owing to the system becoming unstable before the desired sensitivity had been reached. On April 2 the magnet-system was removed from the instrument and