78 MR ROBERT COCKBURN MOSSMAN ON 



February 1879 with a mean humidity of 97, the driest being June 1884 with a humidity 

 of 67. In June 1878 the mean humidity was 69, but in no other month did the mean 

 fall below 70. The month showing the greatest difference between the means is June, 

 the highest being 94 per cent, in 1875, and the lowest 67 per cent, in 1884, the 

 difference being 27, and the month of least difference, December, the highest being 92 

 per cent, in 1876 and the lowest 83 per cent, in 1867, the difference being only 9 per 

 cent. A completely saturated atmosphere is of rare occurrence, not more than two or 

 three cases occurring on an avernge in the year, while in some years no such high value 

 was attained. During the past six years a Eichard hair hygrometer has furnished a 

 continuous record, the lowest value thus registered being 18 per cent, on February 8th, 

 1895. An examination of the hygrograms shows that a humidity below 35 per cent, 

 is of very rare occurrence, even with the shade temperature over 80°. 



Thunderstorms. (See Table XXXVI.) 



During the period 1770-1896, 811 thunderstorms were observed in Edinburgh, 

 or at the rate of six per annum. Of these, 34 took place in winter, 145 in spring, 549 

 in summer, and 83 in autumn. The months of greatest frequency were June with 169, 

 July with 229, and August with 151 ; on the other hand, November and December had 

 only 7 each, while February had 10, and March 11. During the six months, April to Sep- 

 tember, 741 thunderstorms were observed, being 91 per cent, of the whole. Thunderstorms 

 begin to diminish after the Lammas floods, few being observed after the 1 3th of August. 

 The absolute minimum covered the nineteen days ending with December 5th without a 

 single thunderstorm during the 127 years. Lightning without thunder is comparatively 

 rare ; the average annual number of days being only one. Sheet lightning rises to a 

 maximum in September, there being 18 cases in that month during the period under 

 review. A secondary maximum occurs in December. The winter thunderstorms and 

 other electrical phenomena are no doubt associated with deep cyclonic systems ; — the 

 explanation being that in the winter months, ' warm, moist, ascending, and cold, dry 

 descending currents are most frequently brought into close proximity during the great 

 Atlantic storms of the season.' * The diurnal distribution of thunderstorms is well marked 

 (see Table XXXVII.), 64 per cent, being observed during the six hours ending with 

 5 p.m.; the maximum taking place in the two hours ending 3 p.m.; and the minimum in 

 the early morning hours. Lightning without thunder, on the other hand, is essentially a 

 nocturnal phenomenon, nearly all the cases taking place in the five hours ending with 

 11 p.m. Thunderstorms appear to diminish at 1 p.m.; this being doubtless due to the 

 loose way in which certain observers use the word noon. Entries of thunderstorms at noon 

 have all been put down as having occurred in the hour ending noon, whereas half of such 

 entries should have been entered to one o'clock. It was not until the investigation 

 was completed that this anomalous result presented itself. 



The mean annual number of thunderstorms, as already remarked, is six, the year with 



* Ency. Brit., Art. ' Meteorology,' Buehan. 



A. 



