513 



between those dates and the dates of the earliest records is due to there having been no sunshine to record, 

 or to the recorder not having been in use. 



The approximate duration of the sunshine, as recorded on the exposed cards, is shown in the tables on 

 pp. 92, &c. by a thick vertical black line extending across the lines for the hours during which a recording 

 burn was made. 



The following table gives the total amounts of sunshine registered in each of the 14 months covered by 

 the observations : 



TABLE I. Total Number of Hours of Bright Sunshine Recorded in each of the Fourteen Months during 



which, or part of which, the Recorder was in use. 



* In calculating these percentages an allowance was made for those days for which no card was received. 



There are two sets of records for each of the summer months of November, December, and January, 

 during the whole of which the sun was continuously above the horizon ; and for the first summer the three 

 months' records combined give an average percentage of the greatest possible duration of 48 per cent., and 

 for the second year an average of 45 per cent. December was in both years the most sunny month, the 

 total of 376 hours recorded in 1902 being equivalent to 51 per cent, of the greatest possible amount, whilst 

 the more remarkable total of 490 hours in the December of the following year is equal to 66 per cent, of 

 the possible amount. In December, 1903, there was but one entirely sunless day, and the average daily 

 duration of sunshine was 15 8 hours. In January of that year there was also only one quite sunless day, 

 but the average daily duration in that month was four hours less than that reached in December. 



In the summer months there were several long periods of continuous sunshine. In December, 1902, 

 there were three such periods lasting for 38 hours, 41 hours, and 87 hours respectively. In January, 1903, 

 there were two similar spells of 60 hours and of 66 hours ; November of the same year also had two sunny 

 periods of 36 hours and 43 hours respectively; and the following month, December, had three long 

 intervals of 28 hours, 63 hours, and 70 hours respectively. In that month, too, there occurred a remark- 

 ably long spell of almost continuous sunshine which lasted twelve days, from the 6th to the 17th inclusive, 

 during which period the sum of the sunless intervals amounted to less than 15 hours. 



As a rule, the sunshine appears to have been distributed over the months in a fairly uniform way, but a 

 marked exception to the rule occurred in December, 1902, when the total record throughout nine 

 consecutive days in the middle of the month only amounted to 20 hours. 



The hourly totals for the six months November, December, and January, 1902-3, and 1903-4, were 



combined and moaned, and the means smoothed by the formula - , to see if they afforded any 



3 u 



