OBSERVATIONS TAKEN IN INDIA. 
327 
instead of a higher has a lower mean temperature than Calcutta, Bombay or Madras. 
This mean temperature at Madras and Bombay is deduced from hourly observations ; 
and had Dove’s previously noticed mean temperature been derived from similar re- 
cords, the anomalies might have been modified or have disappeared altogether. With 
the reservations contingent upon the explanations given, I shall comprise my obser- 
vations upon the temperature tables of places other than those for Madras and Bom- 
bay, within very narrow limits. And first, with respect to Calcutta, the means are 
derived from observations taken at the Surveyor-General’s office, with excellent in- 
struments, but the observations are day observations, and it is only for the last six 
months of 1848 that records were made every two hours. The means, however, of 
those six months correspond sufficiently near with those of the same six months in 
the three preceding years to show that the conversion of three-hourly observations 
into two-hourly observations, had very little effect upon the mean results. What was 
wanting were observations at night, and of these there are not any. The mean tem- 
perature of the years, from 1845 to 1848, both inclusive, never falls below 83°‘26 (in 
1847), and was as high as 84°* 1 1 in 1845, and the mean temperature of the four years 
is 83°‘72. The coldest month was December, excepting in 1848, when January was 
the coldest. The hottest month was May, excepting in 1846, when April was the 
hottest; the mean temperature of April and May, for the four years, being respect- 
ively 87°' 12 and 90°'27. For the years 1846 and 1847 the mean monthly increment 
to the maximum heat, and mean monthly decrement to the minimum heat, is gradual 
and regular, but in 1845 and 1847 there are instances of inversion; in 1845 March 
is hotter than April; September is hotter than August; in 1848 September is 
also hotter than August, instead of being cooler. This very high mean temperature 
of Calcutta is not of ready explanation, even after making an allowance for the want 
of night observations: situated on a broad river, in the midst of cultivated, well- 
wooded and moist plains, and within the influence of the sea; on the verge of the 
northern tropic, it might have been supposed that its position, in respect to lati- 
tude, would have affected its mean temperature. The sun is vertical at Calcutta in 
the first week in June and July, but Calcutta attains its maximum heat in May, while 
the sun is yet approaching, and the mean temperature is actually diminishing, while 
the sun is passing and repassing from the tropic, and while the heat ought to be ac- 
cumulating from the lengthened days, as is shown in the following tabular statement: — 
Latitude N. 
Longitude E. 
Sun vertical. 
Longest day. 
Calcutta 22 34 40 
Bombay 18 55 42 
Madras 13 4 10 
Aden 12 46 26 
o f 
88 28 15 
72 54 24 
80 21 35 
45 15 0 
June 5 — July 7- 
May 15 — July 28. 
April 25 — August 18. 
h m 
13 23 
13 08 
12 46 
Admitting that the lengthened time for which the sun is nearly vertical over Cal- 
cutta might raise the mean temperature of May, June and July, there would be a 
corresponding diminution in the months of December, January and February, in the 
long nights, and the mean temperature of the year should not be raised by the 
