277
2 2 p. and the coldest about 6 a., but the lowest temperature occurs a little earlier in summer than in winter. When the wind calms down after sunset the existence of a secondary maximum of temperature may occasionally be traced.
The daily range of temperature is smaller in summer than in winter both owing to the greater amount of clouds and to the circumstance that the force of the wind is so small during the night in the former season. From table XVIII it is seen that the mean range was exactly the same at the Observatory as at the Peak,
The monthly extremes of temperature are exhibited in Table XVI and XVII. The ranges are about twice as great in winter as in summer.
The mean diurnal variability of temperature or the mean of the changes of mean daily temperature from day to day, irrespective of sign, during each month of 1885 was as follows: January, 20.37, February, 20.10, March, 29.32, April, 2°.30, May, 1°.44, June, 1°.14, July, 1o.21, August, 1°.25, September, 0°.93, October, 1°.28, November, 20.18, December, 2°.23. Year, 1885, 1o.73.
The mean diurnal variability of temperature in 1886 was as follows: January, 1°.85, February, 20.15, March, 2°.86, April, 1°.63, May, 1°.63, June, 1°.59, July, 0°.98, August, 0°.92, September, 0°.81, October, 1°.13, November, 1°.49, December, 1°.64. Year, 1886, 1°.56.
It is seen from these figures that the temperature is twice as changeable in winter as in summer. Such absence of abrupt changes of temperature indicates in a temperate or arctic climate circumstances favourable to the health of the inhabitants but according to Dr. BORIUS, the late distinguished climat- ologist, great constancy of temperature during the tropical summer may be considered the reverse of favourable.
Table III exhibits the relative humidity in percentage of saturation at the Observatory and at the Peak, the latter being the mean of the 10 a., 4 p. and 10 p. values, and the excess of the hourly values above the mean at the Observatory. The air is farthest from saturation about 1 p. and nearest saturation about midnight. The average relative humidity is registered at 8 a. and 6p. The diurnal variation is greatest during the last month of the year when the air is comparatively dry.
Table IV exhibits the tension of vapour in inches of mercury at the Observatory and at the Peak, the latter being the mean of observations made at 10 a., 4 p. and 10 p., and the excess of the hourly values above the mean at the Observatory. The daily variation is small along the coast of China especially in summer but during the past year it was much greater than in 1885 and this was no doubt owing to the scarcity of rain. There was least vapour in the air at about 11a. and most at 10 p. The monthly ranges in vapour tension (at 10 a., 4 p. and 10 p. only) are exhibited in Tables XVI and XVII. They were twice as great in winter as in summer.
Table V exhibits the total number of hours of bright sunshine. The minimum occurred in February and the maximum in November. The duration of sunshine in percentage of possible duration was obtained as explained in last year's report.
Table VI exhibits the total hourly rainfall and Table VII, the number of hours during which rain was registered. It rains more often at sunrise than at sunset.
The approximate hourly intensity of the rainfall i.e. the hourly rainfall divided with the number of hours, during portion of which it rained, or heavy dew fell, was calculated from the Means and Totals in Tables VI and VII and shows, that the rain is heaviest shortly after noon and lightest about midnight :-
1 a. 0:048
7 a. 0.065
1 p. 0.121
7
2.
.072
8
.071
2
.121
""
p. 0.038 8 .134.
"
3 .059
9
.100
3
.205
9
.169
""
4 .079
10
.105
4
.143
10
.064
>>
**
""
.073
11
.112
5 .142
11
.056
19
6
.067
Noon .098
6
.057
Midt. .071
35
The true mean hourly intensity was obtained from the data in Table XI and shows, that the rain was heaviest in July and lightest in. February and November:-
January, 0.022, February, 0.012, March, 0.036, April, 0.057, May, 0.042, June, 0.073, July, 0.174, August, 0.144, September, 0.136, October, 0.108, November, 0.012, December, 0.033. Year,
0.074.
The rain was much lighter in the past year than in the two previous years, the rainfall being much smaller while its duration was about the same,
Table VIII exhibits the velocity of the wind expressed in miles per hour and the excess of the hourly values above the mean at the Observatory. The velocity at the Peak is computed from the force estimated there at 10 a., 4 p. and 10 p. The wind was strongest at both stations in February and lightest in September. The daily variation was explained in last year's report.
The
Table IX exhibits the mean direction of the wind at the Observatory and at the Peak. excess of the hourly direction at the Observatory above the mean, expressed in degrees, is counted from North through East towards South. The results of the discussion of previous years' observations are confirmed. The daily variation was greatest in September. The two mean directions have been