THE METEOROLOGY OF EDINBURGH. 



471 



October. Entering the number of days during which partial droughts prevailed in the 

 different months we obtain the following values : — 





Jan. 



Feb. 



Mar. 



April. 



May. 



June. 



July. 



Aug. 



Sept. 



Oct. 



Nov. 



Dec. 



Year. 



Days, 

 Per cent., 



26 



2 



98 

 9 



118 

 11 



177 

 17 



85 



8 



243 

 23 



139 

 13 



83 

 8 



9 

 1 



4 







55 

 5 



37 

 3 



1074 

 100 



In the construction of the above table, a drought beginning, say, on June 15 and 

 terminating on July 20 would have 16 days entered to June and 20 to July. The 

 complete absence of spells of dry weather in September and October indicates how 

 unsettled the general weather of these months is. In October the frequency of rainy 

 days is at a maximum, a result of a rapid autumnal fall of temperature and the com- 

 parative absence of high-pressure systems. 



Decennial Means of Solar Radiation for each Day in the Year. 

 {Tables XV. and XVI.) 



The data utilised in this discussion have been derived from daily observations made 

 during the ten years ending with 1898, at my station in the Southside of Edinburgh, 

 254 feet above mean sea-level. The instrument employed was a self-registering solar 

 radiation thermometer, black bulb in vacuo freely exposed to the sun's rays at all 

 seasons, and mounted on a wooden post at a height of 4 feet over a grass plot. The 

 method adopted was to compare the readings of this instrument day by day during the 

 period with the corresponding readings of a maximum thermometer placed in a Steven- 

 son's screen, and enter the excess of the black bulb readings on a form ruled with 365 

 columns, one for each day in the year. If the shade maximum, for example, was 70° *6 

 and the black bulb maximum 100° "2, the difference, viz. 29° '6, would represent the 

 solar excess for the day. What the black bulb in vacuo really records is open to doubt, 

 as no two instruments exposed under similar conditions register the same, the tempera- 

 ture recorded varying with the height above the ground, the perfection of the vacuum, 

 ■ the density of the glass, and the thickness of the superficial film of lamp-black covering 

 the bulb of the thermometer. All that is claimed for the data discussed is, that they 

 have been derived from instruments exposed as described during the period under review, 

 so that the records are strictly comparable inter se. The mean annual excess of the 

 freely exposed black bulb instrument over that placed in the screen is 34 o- 0, the maxi- 

 mum being 48°"9 in May and the minimum 7°"6 in December. The most rapid increase 

 in solar radiation takes place in February and March, and the most marked decrease in 

 November and December. The greatest excess for any day in the year is 55°*5 on May 

 27, and the least 5°'l on December 18. The maximum excess thus occurs nearly a 



