PROFESSOR FORBES ON THE TEMPERATURE OF THE EARTH 



211 



Table XIII. Epochs op Minimum and Maximum Temperature (by Interpolation) 



FROM 1837 TO 1842. 



It will readily be understood, by the inspection of the curves, that these de- 

 terminations are liable to considerable uncertainties, — in most cases amountina: 

 to several days. The curves at small depths are liable to many anomalous fluc- 

 tuations, and even occasionally present an appearance of two minima ; and at 

 great depths the curves, though even, are so flat, that a considerable error may 

 occur in detecting their highest and lowest points. It does not appear, however, 

 that more real accuracy would be obtained by the methods of calculation which 

 have usually been employed, instead of interpolating curves. We shall presently, 

 however, shew how the two may be advantageously combined. 



We thus see that the greatest cold of winter attains the depth of 24 French 

 feet, — 



At the Observatory (trap rock), on the 13th July ; 



At the Experimental Garden (loose sand), on the 29th June ; 



At Craigleith (sandstone), on the 3d May ; 



and that the greatest heat occurs on the 4th January, 25th December, 3d Novem- 

 ber respectively ; shewing, in both cases, the very same order of facility in con- 

 ducting heat which we had before deduced from the diminution of ranges, namely, 

 that the Observatory ground is the worst conductor, that of the Experimental 

 Garden but little better, and the rock at Craigleith by far the best. 



Unfortunately, the measure of the retardation of epochs has, as yet, been so 



