IV. CLIMATE. 155 



cestersliire, a portion of the same province of Severn, 

 But implicit reliance cannot usually be placed on the 

 high and equable temperatures which are recorded for 

 places of invalid resort, by medical practitioners and 

 others interested in publishing the most attractive ac- 

 counts of theii' local climates. The mean of 60 may, 

 however, be near truth for the coasts of the Bristol 

 Channel, decreasing inland and northward. 



Looking to North Britain, records show a mean of 47, 

 with a fi-action under or a fraction over that figure, for 

 low situations in the most southerly portions, that is, in 

 the sub-province of the South-east Highlands. Those 

 for the town of Inverness, by registers in different years, 

 show a remarkable discrepancy ; being formerly stated at 

 48|, and more recently at 451-. Taking the mean be- 

 tween these two, it would be 47, which seems more than 

 sufficiently high for that northern locality, with the ele- 

 vated ranges of the Grampian hills to the southward, and 

 other lofty hills to the westward also. Judging either by 

 the geographical and topographical position, or by the 

 actual vegetation, the temperature on the north side of 

 the Grampian hills should certainly be below that on the 

 south side. So that, reverting to the two figures above 

 mentioned, it may be surmised that the mean of 45^^ will 

 be something nearer to the true average than that of 48|. 



Northward from Inverness the records become too 

 short and uncertain for much trust. For Wick, on the 

 coast of Caithness, means of 45|- and very slightly under 

 47 have been reported, say 46^. At Stromness, in 

 Orkney, the mean is stated to be rather higher than 46;^ ; 

 and exactly the same figure is recorded for Sandwick, in 

 the same isles. The hypothetically corrected mean for 

 Unst, in Shetland, would appear to be under 45 of 

 Fahrenheit (see vol. 2, page 4). Likely enough, the 



