76 



MR ROBERT COCKBURN MOSSMAN ON 



on observers to have rain-gauges capable of holding at least 4 inches of rain, otherwise 

 important facts of interest to meteorologists and engineers alike will be irretrievably 

 lost. The following are the maximum daily rainfalls noted in each month during the 

 period under review. 



Month. 



Jan. Feb. 



Mar. 



Apr. 



May. 



June. 



July. 



Aug. 



Sept. 



Oct. 



Nov. 



Dee. 



Rainfall, 

 Date, . 



1-59 

 10/1809 



1-80 



3/1809 



1-55 



16/1891 



171 



5/1808 



1-50 



14/1795 



1-54 



26/1874 



2-95 

 13/1879 



2-56 



18/1797 



3-80 



24/1785 



2-50 

 5/1775 



2-89 

 18/1795 



4-20 

 9/1787 



Direction of the Wind. 



Table XXXII. shows the number of days on which each wind prevailed, from June 

 1731 to May 1736, and from 1764 to 1896 ; for the months and the year. From 1764 

 to 1769 the only values available are the summaries of east and west wind prepared by 

 Hoy. The former includes observations from N., N.E., E., and S.E., the latter those 

 from S., S.W., W., and N.W. As 50 per cent, of the winds in Edinburgh are from the 

 8.W. and W., and 25 per cent, from N.E. and E., it follows that the above method of 

 reducing the wind observations to the two principal directions gives a close approxima- 

 tion to the truth. The registers employed are those utilised in the preparation of 

 daily values.* From 1781 till the commencement of Waterston's observations 

 in their complete form in 1805, the direction of the wind was not systematically 

 observed. It was therefore necessary to interpolate from the Glendoich and Dunfermline 

 registers, which in some measure help to supply the deficiency. Table XXXIII. shows 

 the mean percentage frequency of the winds for the months and the year for the 133 

 years 1764-1896. The mean values for 100 years are given in Part I., and are herewith 

 compared with the longer record. 









Percentage Fi 



'equency 













N. 



N.E. 



E. 



S.E. 



s. 



S.W. 



W. 



N.W. 



Calm or W. 



133 Years, . 

 100J „ 



4 

 4 



7 

 7 



18 

 16 



5 



7 



5 



6 



15 

 17 



35 

 32 



7 

 7 



4 

 4 



The means for the two periods are thus essentially the same. The observations were 

 made twice a day during nearly the whole of the 133 years. 



In their reduction the values were resolved to eight points by counting N.N.E., for 

 example, along with N.; S.S.E. along with S.; E.N.E. along with E., and so on. This 

 was done in order to make the observations taken prior to 1856 comparable with those 

 irivn in the Scottish Meteorological Society's Journal during the last forty years. 



* Trans. Roy. Soc. Edin. t vol. xxxviii. p. 691. 



