Probable Error of the Time Ball.
Month. 1918 1919 1920 1921 1922 January, 0'11 ±0·24 February, *20 +0°17 *30 ±0.25 ±0.10 13 15 March... 15 '12 '21 44 *12 April, '10 19 15 2 *20 May, '12 *14 17 16 *IO June, 14 14 13 17 'I I July, II *13 •22 • August, *10 14 26 15 September, *I I 10 *24 *20 15 15 October, '12 15 15 *10 *10 ► November, 12 14 19 *10 *17 *10 December, 14 *12 *13 Means, ±0.14 ±0.15 ±0.18 +0.18 ±0.13Time Signals by Radio Telegraphy. In addition to the time signals given by the Time Ball, and on the radio mast, signals are sent at 10h. and 21h. by radio telegraphy via Stonecutters. Particulars of the programme are given in the 1918 Report and in Government Notification No. 452 of 12.3.21. The service was transferred from Cape d'Aguilar to Stonecutters on May 1, 1921.
Radio Receiving Set. The radio receiving set was in irregular use throughout the year. 109 comparisons were obtained with the Manila Observatory clock via Cavite, and 27 with the Tokio Observatory clock, via Funabashi.
The mean of the comparisons makes Tokio 0·54 sec. fast and Manila 10.3 secs. fast on Hongkong.
Between July 28th and August 26th Manila was not heard although listened for daily. This may have been due to defects in the receiving apparatus, the adjustments of which are extremely critical. From November 29th to December 26th no signals were heard, in spite of repeated attempts to adjust the apparatus. It was thoroughly overhauled and re-wired on December 22nd and 23rd since when it has worked satisfactorily.
The Sicawei time signals (via Koukaza) were not heard, although listened for on many occasions.
Transit Instrument.-Observations for time were made daily with the 3-inch transit instrument and the Hipp tape chronograph by the Chinese computers, weather permitting.
The number of observations in the years 1921 and 1922 were as follows:-
1921 1922 Transits 1,502 1,307 Level determination 869 696 Azimuth.... 50 31 Collimation... 37 28- Fin