THE EARTH IN METEORIC SHADOW. 143 



would be found tolerably uniform. Thus for London 

 and its neighbourhood we might expect to find some- 

 thing like that variation of temperature which is in- 

 dicated in our almanacs (which leave altogether out of 

 account, for some unknown reason, the anomalies we 

 are considering). So that if we represented the average 

 temperature for successive days by an upright line, 

 drawn from a horizontal line indicating the positions of 

 successive days in the year or, to speak technically, if 

 we represented daily average temperatures by a series 

 of ordinates to a line along which time was indicated 

 by abscissas we might expect the curve passing 

 through the upper extremities of the uprights to have 

 a wave-like form, the crest of the wave lying above the 

 part of the horizontal line corresponding to the middle 

 of July, while the hollow or valley of the wave lay 

 above the part corresponding to the middle of January. 

 But this is not found to be the case. I have before 

 me as I write a diagram drawn by myself several years 

 ago in illustration of my article on the Climate of Great 

 Britain in the second series of my < Light Science for 

 Leisure Hours.' I drew a large rectangle, and divided 

 each of its longer sides into 365 parts, to represent the 

 days of the year, and drew through the points of 

 division a series of 365 uprights, on which I marked 

 the mean annual temperatures for the corresponding 

 days, the mean having been derived from Greenwich 

 observations ranging over forty-three years. A con- 

 nected line was carried then through the 365 extremi- 

 ties of these lines. 



