46 Register of the Weather at Futtehgurh. [Jan. 



article has been lost. They remain under water from one to one and a 

 half minute at a time — oft-times more, if the water be deep. They 

 adopt the same system precisely, whether in still water or in a running 

 stream : only that in the latter, of course their labour is more severe — 

 their success more precarious. 



Their remuneration depends solely on success ; the ordinary salary 

 being one-third of the extricated value of the lost article, and which 

 is divided in equal portions among the set. 



X.— Register of the Weather at Futtehgurh (Lat. 27°21'iV. Long. 79°30' 

 E.)from April 1832 to October 1833. By M. P. Edgeworth, Esq. C. S. 

 The thermometer was placed in the open air, on a wall fronting the 

 north, until 1st January, 1833, when it was removed to an open veranda 

 on the north side of the house. Up to 26 Sept. the maximum was taken 

 by a self- registering thermometer, which was accidentally broken : it was 

 then taken at 2^hp. m. till December, "1 



2 p. m. till April, V by a spirit thermometer, 



3 p. m. till August J 



and from August 6, by a self-registering thermometer. The minimum 

 all along by a self- registering spirit thermometer. 



Note. We have endeavoured to render the abstract, into which want of 

 *pace has obliged us to condense our correspondent's register, more com- 

 plete by expressly numerically the number of days, windy, cloudy, fair, &c. 

 in each month, as far as can be gathered from a register not intended to 

 shew these points with accuracy. The columns of west and east wind 

 comprehend 45° degrees on either side of the cardinal point, as it seemed 

 more proper to class these winds (north-west, south-east &c.) with the 

 directions generally prevalent, than with the north and south winds, 

 which are of rare occurrence. 



The mean temperature of Futtehgurh seems nearly as high as that of 

 Benares or Ghazipoor*, but we are not aware that the instruments used 

 had been previously compared with a standard. 



For four days of 1832, Mr. Edgeworth took the temperature every 

 hour during the day and night : which enables us to prove that the sup- 

 position of deriving the mean temperature of a place from the means of 

 • two hours of the same name will not hold good. At the foot of the hourly 

 register we have given the means of the pairs thus deduced; and 

 under them the errors from the mean of the whole (75°. 55), which may 

 be taken as the corrections due to each pair. The mean of the extremes 

 of heat and cold (7 6°. 55) is 1.00 higher than the mean diurnal range. 

 In my register for Benares (App. x. As. Res. xv.) I found the excess to 

 be 0.86, which is a near accordance with Mr. Edgeworth's result. — J.P. 

 * See vol. i. 29, and vol. ii. 604. 



