63 



June the mean depression of the dew point below the dry- 

 bulb thermometer was more than 1 degrees ; while the 

 relative humidity of the air (saturation being 100) in May 

 was 69, in June 72, some days in the latter month as 

 low as 56. In July the mean depression of the dew point 

 was not quite so great, but the northerly winds were more 

 constant. 



There had been two thunderstorms previously, in April ; 

 one on the 21st, the other on the 25th; but the weather 

 during the latter half of that month was very unsettled, with 

 an excess of rain, and the air generally in a much more 

 humid state than in any of the three following months. All 

 this seems to afford testimony to what has been above stated 

 with respect to the conditions under which thunderstorms 

 usually take place. 



It may be added that the summer of 1835, which was very 

 hot and dry, like that of 1868, was also characterised by the 

 occurrence of very few thunderstorms. 



Having now spoken of the principal features observable in 

 the summer of 1868, I proceed to notice a question which 

 some may be disposed to ask, and that is, whether we are 

 ever hkely to be able to predict the character of the coming 

 season, so as to foresee a hot summer like that we have just 

 experienced ? It would be presumptuous confidently to assert 

 what science, even so backward a science as meteorology, 

 compared with some others, may, or may not, be equal to in 

 years to come. Assuredly, however, we are not in a position 

 to do this at present. We may note some things as favour- 

 ing our expectations. But while even the ordinary causes 

 that combine to influence weather are so many and so com- 

 pHcated as not easily to be disentangled, there arise from 

 time to time contingencies, which no human sagacity can 

 foresee, to confound all our calculations. 



I have no notes by me to refer to respecting all the 



