72 MR ROBERT COCKBURN MOSSMAN ON 



The mean annual variability of temperature is2°"85, being highest, 3°'24, in January 

 and lowest, 2°"52, in July, thus showing a difference of 0°72. The greatest variability 

 was 3 0, 38 in 1843, and the least 2 c- 50 in 1860, the range in the annual means being less 

 than 1 degree. The greatest variability of any month was 4° - 9 for November 1847, 

 while the low value of 1°'G was recorded in the Julys of 1853 and 1854, the Augusts 

 of 1858 and 1860, and in September 1861. The greatest daily rise of temperature 

 occurred on March 17, 1892, whose mean temperature was 15°'l higher than that of the 

 16. August 29, 1869, on the other hand, was 15 0, 5 colder than the previous day. The 

 daily observations for fifty-seven years were gone over, each rise or fall of 10° or more 

 in the mean temperatures being extracted. The number of such cases was 230, viz., 

 129 rises and 101 falls (see Table XXVI.). The greatest number was 14 in 1843, and 

 the least 1 in 1857, 1859, 1861, 1862, 1883, and 1891. In six of the years there was 

 no fall of 10°, and in four of the years no rise of 10°. The greatest number of 10° rises 

 was in 1843 and 1845, when nine cases were recorded, while the maximum number 

 of 10° falls, viz., six, occurred in 1880. As the variability of temperature at stations 

 on the Continent is as a rule calculated from observations taken at stated hours, and not 

 from the mean of the maximum and minimum, Table XXVII. has been prepared. This 

 Table gives the mean daily temperature variability for the hours of 9 a.m. and 9 p.m. 

 which are then compared with the values deduced by taking the daily means of the 

 maximum and minimum. Table XXVIII. shows the means deduced from the 8 a.m. 

 observations taken by Hoy at Hawkhill House, and Kirkcaldy, while corresponding values 

 for the period 1731 to 1736 are discussed in another section. It has been shown that the 

 variability of temperature is subject to a diurnal range,* but unfortunately the Edinburgh 

 records are sadly defective in data from which hourly values could be calculated for 

 this or any other climatic element, with the single exception of sunshine. 



Rainfall. 



Table XXIX. shows the monthly and annual rainfall in Edinburgh for 120 years 

 and six months. The values from 1770-76 were taken by Hoy at Hawkhill. Mr Hoy 

 was also the observer during 1780 and the first half of 1781 when he removed to Gordon 

 Castle. From 1785 to 1794 the observations were deduced from the Edinburgh 

 Magazine record, the gauge being at Duddingston till January 1793, and thereafter 

 " within one mile of the Castle." The values from 1795 to 1805 and from 1822 to 1850 

 are those taken by Mr Adie, and given by Forbes in his Climate of Edinburgh. The 

 late Mr Leslie commenced his long series of rainfall observations in 1850, the station 

 being Charlotte Square, where the record is still continued. The returns from this 

 station have been utilised for the period 1851-96. 



From 1805 to 1821 rainfall was not systematically observed at any one station 

 during the whole period ; but values have been obtained from measurements made 

 at the Royal Observatory, and at other places in Edinburgh. I am indebted 



* Jour. Scot. Met. Soc, vol. x. p. 150. 



