Mr. F. D. Brown's Notes on Thermometry. 65 



during the next two or three hours the rise was extremely 

 small. It follows, therefore, that if in any series of observa- 

 tions lasting more than three hours the thermometer be heated 

 in steam at the end of every third hour, there will be no un- 

 certainty as to the position of the zero; that if the experiments 

 be carried on continuously for six hours, a slight rise of the 

 zero may occur during the last part of the time, but that this 

 rise will not amount to more than one or two hundredths of a 

 degree. 



Correction for the Exposed Portion of the Thread. 



When a thermometer is only partially immersed in the 

 medium of which the temperature is to be observed, the 

 readings become subject to an error which arises from the 

 fact that a part of the thread of mercury, together with the 

 corresponding portion of the stem, are at a temperature 

 different from that of the bulb and immersed portion of the 

 stem. The correction, C ; usually applied in this case is given 

 by the formula 



C = m(T-0N, (1) 



where T=the reading of the thermometer, 



f = the temperature of the exposed portion, 

 N=the number of exposed divisions of the stem which 



are filled with mercury, 

 m = the apparent expansion of mercury in glass. 



This formula is founded on the assumption that the error in 

 the reading has no other cause than the comparatively unex- 

 panded condition of a portion of the thread and stem. 



The apparent expansion of mercury in glass, as obtained 

 from Eegnault's experiments, is about '0001545; but it differs, 

 of course, for different specimens of glass. When this number 

 is employed in the above formula, the values of C obtained are 

 generally believed to be too large ; indeed a little reflection 

 will convince us that this must be the case whenever the tem- 

 perature of the exposed portion is merely measured by placing 

 another thermometer with its bulb halfway up it. This second 

 thermometer evidently measures the temperature of the 

 ascending stream of warm air around the stem ; if the stem 

 of the chief thermometer were subjected to the heating 

 influence of this stream, and to no other, its temperature 

 would be rightly given by the subsidiary thermometer ; but 

 the thermal conduction along the thread of mercury and along 

 the glass stem must necessarily raise the lower part of the 



Phil, Mag. S. 5. Yol. 14. No. 85. July 1882. F 



