H. F. Reid — Theory of the Bolometer. 161 



If d is the deflection of the galvanometer needle, and D a 

 constant depending on the form of the galvanometer, the 

 period of the needle, etc., for small deflections 



G(l+l)+l(R + Sy 



A particular bolometer will have its greatest sensitiveness 

 when for a given value of k—l, and a given 'percentage probable 

 error of observation, the deflection, d, is greatest. This is ob- 

 tained by giving proper values to D, G and v. The proper 

 values of I) and G must be determined independently of the 

 intensity of the current, v ; for changes in these values will 

 merely alter the sensitiveness of the galvanometer, and will 

 change in the same ratio the total deflection d, and the irregu- 

 lar deflections of the needle, but will not change the percentage 

 probable error of observation. D should be made as large as 

 possible ; i. e. we should select the best form of galvanometer, 

 have the needle strongly magnetised and highly astatic and its 

 period long. The best value to give G is l(R + S)/(l+l). The 

 current v must be increased to its greatest practical intensity ; 

 this limit, which is not at all well defined, is reached when the 

 strip becomes sufficiently heated to set up irregular air cur- 

 rents, which cause fluctuations in its temperature and irregular 

 movements of the galvanometer needle, thus increasing the 

 probable error of observation. 



Introducing the above value of G in the equation and writ- 

 ing S=nR we get 



s=v «y R (*-*) 



2V(l+n). 1(1+1) 

 We have supposed the arm Mi to contain the exposable strip 

 of the bolometer. When this is screened from radiation and 

 the bridge properly balanced, h=l, and the resistance of this 

 arm is ZR ; when exposed to radiation let its resistance kJH=qlR ; 

 then 



s= — D zLi)(g-i)yl. (i) 



2V{l+n)(l+l) W 



IR is the resistance of the bolometer strip, and the wires 

 connecting it to the bridge, when unexposed, and is of course 

 a fixed quantity for any particular instrument, d increases as 

 I and n decrease. In order that the balance of the bridge 

 should not be destroyed by the continued variation of the tem- 

 perature of the room, it is found important to have the arms 

 of the bridge, two and two, as nearly alike as possible ; a 

 second arm therefore is made of a platinum strip like the first 

 and placed near it in the same case, but entirely covered up. 



Am. Jour. Sci.- Third Series, Vol. XXXV, No. 20G.— Feb., 1888. 

 10 



