430 



Transactions of the South African Philosophical Society. 



The values of X found above are not necessarily the actual working 

 coefficients of these thermometers, the conditions under which they 

 were determined differing entirely from those to which they are 

 exposed in meteorological investigations ; but they serve the purpose 

 of this discussion in giving approximately comparative values for the 

 different thermometers employed. 



It would appear, moreover, from the persistent decrease in the 

 computed value of X, that for some thermometers at least, with the 

 value of (t — t) employed, the relation between t, r, X, is not correctly 

 given by the equation — 



dt 



-)=0; 



but at the same time it should be remembered that the generally 

 large value of X found for the interval = 6 sees., depends on a 

 portion of the curve where t is decreasing rapidly, and where, in 

 consequence, an error of a few tenths of a second in will give large 

 errors of t. 



Curves, derived from the data given in Table IV., indicating the 

 degree of sluggishness of the several types of thermometers 

 employed are given in Fig. 3, 



■ 

 Jah 



Seconds 

 10 2 



7 20 10 20 30 40 10 2 



7 30 40 SO 



10 20 30 40 SO 60 70 80 90 100 110 





\ 

 Standard minimum Spirit 



| 























78 





JV7/4 





Jype J 

 N° 



rr 



75616 





Jype 



1 JY' 



II 

 II55II 





J 



Jype 

 .A" 



I 



147 O 































\ 

























































































-70- 











































































































Fig. 3. 



In the following tables will be found the details of the construction 

 of each thermometer, together with the errors as determined at Kew 

 and at the Eoyal Alfred Observatory, Mauritius. 



