IV. CLIMATE. 168 



be made to deduce from the mingled mass some sort of 

 rules or formulas, which may be more readily recollected 

 by their uniformity, and thus become more applicable to 

 botanical purposes ; although any forced uniformity will 

 be a deviation to some extent from at least a portion of 

 the actual registers. 



13. Relation of Mean Temperature to Latitude. — Pub- 

 lished records mostly agree in showing 47 or 48 of 

 Fahrenheit's scale, with some fraction or decimal added, 

 for the mean annual temperature of places in Mid Britain, 

 not raised much above the level of the sea. But local 

 discrepancies are here too numerous and conflicting, 

 ■without really being very wdde, to allow of any rate of 

 valuation being quite satisfactoril}' fixed, so as to connect 

 the temperature with the latitude in an uniform manner. 

 It would be possible to select one series showing a 

 decrease of mean temperature from south to north, — 

 another series seeming to prove the contraiy, or an 

 increase from south to north, — and a third series that 

 would half contradict both, by its alternations of increase 

 and decrease. Still, the prevailing character seems to be 

 a slight decrease of mean temperature in a northerly 

 direction, and a somewhat greater decrease of summer 

 temperature in the same direction. Passing hence into 

 the two other latitudinal divisions, those of South 

 Britain and North Britain, the direction of decrease in 

 temperature becomes clear enough, and a rate of decrease 

 not so difficult to decide upon by way of general rule. 



In South Britain, it might at first thought seem 

 reasonable and safe to look to the registers for London, 

 as kept for the Royal Society during a long series of 

 years, and regularly published up to a late date, appa- 

 rently under the sanction of the Society's officers. But 

 VOL. rv. X 



