48 The Meteorology of 1898. 



regard to tlie monthly means of temperature, the highest was in 

 August, with a mean of GO'4 deg., the mean of July was 59-1 deg., 

 June 57'(5 deg., and September 57-5 deg., so that August was the 

 warmest month, but July had the highest mean maximum, viz., 

 G9"2 deg. as compared with 68-2 deg. in August. It is worth}- of 

 being observed that no less than nine of the months of 1898 had a 

 temperature in excess of the average, with an aggregate of no 

 less than 27*5 deg., while the deficiencies in the remaining months 

 were so slight as to make a very trivial deduction from this excess. 

 One remarkable feature of this excess was that it occurred chiefly, 

 although not exclusively, in what are usually the coldest mouths — 

 January and December. January, for example, had over 6 deg. 

 above the mean, and December 5-7 deg. (almost 6 deg.), and 

 October over 5 deg. But April, May, August, and Sepi ember had 

 also considerable excesses, ranging from 1-2 to 2-9 deg. This 

 suggests an unusual mildness of the winter months, which comes 

 out very distinctly when we take into account the number of 

 nights on which the protected thermometer fell to and below 

 the freezing point. These were only 38 in all, with an aggregate 

 of 103 deg. of frost. This exhibits quite an exceptional state of 

 things as compai-ed with the average of other years. January 

 and December are usually the coldest months, as might be 

 expected from the shorter period during which the sun is above 

 the horizon, and also the greater obliquity with which the solar 

 rays fall upon the earth. But in 1898 the mean temperature 

 of both these months was no less than 44 deg., only 2 deg. 

 short of the average of April, with only five nights of frost, two 

 in January and three in December, with an aggregate of 21*8 

 degs. ; while February alone had ten nights with an ag'gregate of 

 32 deg., and March eighteen with an aggregate of 41 deg. The 

 two latter months were the coldest of the year, with a mean of 40 

 deg. , as compared with 44 deg. in January and December ; and 

 yet the temperature of February was a little above the mean, and 

 March only slightly below it. The explanation of the high 

 temperature of January and December is to be found in the 

 unusual prevalence of southerly to westerly winds, and the con- 

 vection of heat from the warmer to the colder regions by means of 

 atmospheric currents, which greatly modify the effects of solar 

 radiation. To illustrate this, I may mention that southerly, south- 

 westerly, and westerly winds prevailed for no less than 26 days 



