THE EARTH IN METEORIC SHADOW. 145 



three small undulations, which lie nearly in a horizontal 

 direction, instead of indicating the continued rise we 

 might fairly expect at this season. In the last half of 

 July there is a rise to the highest part of the curve, 

 which then begins to fall rather sharply, the fall being 

 less rapid after August 10 (instead of more rapid as we 

 might fairly expect). Thence, to the last half of 

 November, there is a tolerably steady fall, but then 

 occurs a rise, making the last two or three days of 

 November and the first week of December considerably 

 warmer than they theoretically should be. After the 

 middle of December there is a very rapid fall in the 

 temperature curve, but a rise occurs in the last week 

 of December somewhat similar to that observed in the 

 last week of November. 



The most marked features of the curve are the 

 three cold periods of February, April, and May, and 

 the warm period at the beginning of December. 



Let us turn, however, from the south to the north 

 of Great Britain, and see whether the peculiarities 

 noted at Greenwich exist at Edinburgh. 



Buchan, in his Handy Book of Meteorology, 

 remarks that the results of all observations hitherto 

 made are unanimous in showing that ' there are certain 

 periods, more or less denned, when the temperature, 

 instead of rising, remains stationary, or retrogrades 

 instead of falling, stops in its [downward course, or 

 rises and at other times falls, or rises for a few days, 

 at an accelerated speed. I have examined,' he says, 

 'the temperature of Scotland, for a number of years > 



III. L 



