11. The probable errors of the signal in the different months of 1887 (with the average percentage of clonded sky added in parenthesis) were as follows:-January 0.18 (90), February 0.31 (64), March 0.15 (75). April 0.28 (72), May 0.17 (81), June 0.16 (71), July 0.15 (73), August 0.11 (59), September 0.20 (64), October 09.11 (27), November 0.14 (41), December 0.18 (43).--The mean of the probable errors was 0.18.
12. The absolute values of the magnetic elements were observed by Mr. F. G. FIGG and Mr. MAHOMET ÁLARAKIA every month as usual, and the report is ready.
13. The monthly weather reports have all been published as soon as ready. They are arranged exactly as explained in last year's report. The clouds are classified as follows:-
Cirrus (c) is the finest and most lofty of all the clouds. It looks like hair, thread or feathers. Cirro-cumulus (c-cum) is also a high cloud, but is more rounded in shape than the former and looks like small woolly tufts.
Cirro-stratus (c-str) form a sheet of uniform thickness high up in the atmosphere. They are often seen in long straight streaks, that appear to radiate from a point of the horizon. When covering the whole sky they form a sort of vaporous transparent veil,
Stratus (str) is also a layer of cloud of generally uniform thickness but belongs to the lower regions of the atmosphere. It presents no variety of shade.
Cumulus (cum) consists of rounded heaps like enormous balls of wool.
Cumulo-stratus (cun-str) is a cumulus dark and flat at its base traversed by horizontal streaks of dark cloud. It has often a coppery hue.
Strato-cumulus (str-cum) is intermediate between cumulus and stratus. When the number of entries are counted, half is added to cumulus and the other half to stratus.
Roll-cumulus (R-cum) is formed by cumulus clouds lying apparently in long horizontal rolls.
Small-cumulus (sm-cum) are small white cumulus. They belong to a level between cirro-cumulus and cumulus.
Nimbus (nim) is a cloud from which rain is seen falling.
Cumulo-nimbus (cum-nim) has a uniform dark appearance like the true nimbus, but no rain is actually seen falling from it at the time of observation.
14. The weather in January 1887 was quite abnormal. The rainfall was about six times larger than the average. The amount of clouds and the dampness were also excessive particularly the former, so that the rainy season appeared to have set in in the midst of the dry season and that was the case at every station in China from which reports are received. On the contrary the weather was remarkably dry and clear in March, gradients for NE winds continuing till the 29th of that month. On the 8th of April the damp weather set in, but the rainfall was very deficient in that month and also in May. The 1st of June was extremely dry and the barometric tide and daily variation of temperature excessive for the season as might be expected. The middle of that month was very squally with strong SW monsoon. Showers in the carly morning hours were distinctly marked but electric phenomena were unusually rare. The thunderstorms during July came from unusual directions. In August the SW monsoon was strong, and the clouds unusually low, as was in fact the case all through the summer but still the rainfall was below the average. September was characterised by an unusual number of typhoons in this neighbourhood as remarked at the end of the report for that month. The weather in October was very dry, clear and fine. On the 21st during N wind it was hotter in the puffs, reminding one of the Fühn in Switzerland. December was very dry and warm. At night time there occurred frequently a sudden change in the direction of the wind from about E to about N with considerable rise of temperature and great decrease of humidity.
15. Tables III-X inclusive exhibit the annual means and totals of meteorological phenomena and their frequency. The arrangement of the tables is explained in previous annual weather-reports. Tables XI and XII exhibit the five-day means arranged as in previous years. The total amount of rain measured daily at 10 a. and cutered to preceding day was 66.29 at the Observatory, 61.73 at Stone Cutters' Island and 78.34 at the Peak. The total duration registered at the Observatory was 863 hours. The rainfall was at least 0.01 inch on 153 days at the Observatory, on 112 days at Stone Cutters' Island and on 84 days at the Peak. The accuracy of these observations is in a great measure due to the painstaking care of Mr. F. G. FIGG.