30 



made every two houi-s for rather more than a month, before they 

 were put into the ground, is as follows : — 



o 



No. 1 — 12 feet long, requires the addition of 2.133. 



2 — 6 2.172. 



3 — 3 2.922." 



This detail is even too minute to render a considerable mistake 

 at all likely ; we must therefore take the results as we find them, 

 which are contained in the following table, "where the numbers are 

 all corrected, and compared with the mean temperature of the air, 

 and the quantity of i-ain. It will be obsei'ved, that several of the 

 observations of the warmer months have been lost, owing to the large 

 index error having misled the maker as to the range which would 

 be requu-ed, and consequently the spirit rose above the Scale. 



Abstract of Observations of Terrestrial Temperature at Trevandrum. 

 Lat. 8° 30' 32" N. Long. 5'> 7°' 59s e. of Greenwich. 



The Thermometers are corrected for index error. 



With a view to deduce more carefully the results of this curious, 

 and hitherto unique series of observations, I have had the observa- 

 tions for the year projected in the form of curves, on the same scale 

 as those formerly submitted to the Society, of the earth tempera- 



