PHILOSOPHICAL SOCIETY OP WASHINGTON. 69 



led liim to place another thermometer on the west side of the ob- 

 serving room and make a series of comparisons with the standard. 



Thermometer A was suspended on tlie outside of the northern 

 wall of the observing room in a double tin box shaped like the 

 frustrum of a pyramid with a base 30 inches square. 



Thermometer B was placed on the northern wall near the roof, 

 and just outside the eastern side of the vertical opening in the 

 northern wall of the observing room. 



Thermometer C was attached to a rod depending from the roof 

 almost directly over the transit circle, and was nearly on the same 

 level as the object-glass when the instrument was pointed at the 

 zenith. 



The standard thermometer D was in the meteorological observ- 

 atory, situated southeast of the observing room. 



Thermometer A was 7.5 feet above the floor of the observing room. 

 B " 16.5 

 " C " 13.5 " " " " 



D " 4.2 " the ground on the lawn. 



From A to B the distance was 11.0 feet. 

 " A to C " " 23.5 " 



" B to C " " 21.5 " 



" A to D " " about 37 yards. 



Sixteen comparisons were made between May 8 and July 16, 

 1870, and the most marked differences were as below: — 



From the following equations giving the greate.-^t positive and 

 negative dilFerences the ranges of differences are obtained : — 



A— B = +4^0 Range 4°.0 



A — = + 6°. 5 



= — 3^0 Range 9°. 5 



B — C = +4^2 Range 8^2 

 = — 4^.0 



Representing the mean of the readings of the exterior and 

 interior thermometers, A and C, by M, we have the maximum 

 value of M — D = 6\9, and the minimum value = l°.l. Also 

 the maximum value of M — B = 3^.8. 



Assuming the temperature indicated by the standard thermom- 



