24 S. A. Hill — Observations of the Solar Thermometer. [No. 1, 



rally held by solar physicists, I have always looked upon it as doubtful, 

 and probably due in part to some fortuitous combination of errors. I 

 therefore intend on some future occasion, possibly after the end of the 

 present year, when the position of the thermometer at Allahabad will be 

 changed, to go over the figures again, taking a longer series of obser- 

 vations and making allowance for a cause of variation from month to 

 month, namely, the elliptic form of the earth's orbit, which was neglected 

 in the paper referred to. Meanwhile, I wish to lay before the Society the 

 results of some other observations bearing on the same question, which 

 tend to confirm the conclusions arrived at in my previous paper. To the 

 method by which these results are attained, less exception can be taken, 

 because they are in every case derived from several observations made 

 on the same day under different degrees of obliquity of incidence, in- 

 stead of upon the single record of a self-registering instrument. 



Shortly after hourly observations on four days in each month were 

 commenced at Lucknow, it was discovered that the solar thermometer 

 in use at that station had ceased to be self -registering. A new instru- 

 ment was therefore brought into use on ordinary days, but the old one 

 was retained for the hourly observations. The records of all such obser- 

 vations of this instrument since the middle of the year 1876 have been 

 filed, but for the purposes of the present paper I have used only those 

 of the eight years 1877 — 1884 inclusive. At Agra, similar observations of 

 a non-registering solar thermometer have been made for some years on 

 hourly observation days, but, owing to a change of instrument, the 

 register for the years 1877 — 1884 is broken. For this reason, and because 

 the observatory at Agra is situated in the midst of the city, I have not 

 thought it worth while to reduce the registers of that station, though 

 they seem to confirm in a general way the results obtained from Luck- 

 now. 



Those parts of the Lucknow records which have been used for the 

 purposes of the present paper are printed in Table I. The figures represent 

 for each hour of observation the difference between the temperature of 

 the black-bulb thermometer in the sunshine and the simultaneous tem- 

 perature in the shade. Only those hours are given at which the sky 

 was either quite free from cloud or at which the cloud proportion did 

 not exceed 2-lOths of the expanse. In the months of July and August, 

 very few clear days, thus defined, occur ; consequently these months 

 have been left out in drawing up the tables. For every other month in 

 the eight years, except September 1878 and June 1880, there are some 

 observations available. 



