80 Intelligence and Miscellaneous Articles. 



it is right to remember that in a solar spectrum, especially when it is 

 obtained with glass and Iceland spar, the heat is almost entirely 

 confined to a space which does not occupy one sixteenth part of 

 the total length of the spectrum. The value of n employed in this 

 case is the mean of those corresponding to the different single rays 

 of that narrow region. 



I have thought it proper to notice the accord shown in the pre- 

 ceding Tables ; but I mention it as a fact, without wishing to enter 

 into any theoretic discussion. Equation (1), which stands here in 

 the same form as in my former communication, might quite as well 

 be written sin z cos?- .,, , ., 



sm r cos i 



Equation (3) would then become 



tang (i — ;•) _ _ / i , ^ sin z cos i \ 



tang (z + r) \ sin i cos i + sin r cos r / 



h sin i cos i 



sin z cos / + sin r cos r 

 and the values of K would become positive, but without changing 

 their absolute quantity. I add, in conclusion, that a clerical error 

 crept into the final equation in my last communication ; that error is 

 rectified here. — ComptesRendus de V Acad, des Sciences, April 29, 1872. 



ON ELECTRICAL PYROMETRY. BY LIEUT. ABXEY^ R.E.^ E.R.A.S., 

 P.C.S._, ASSISTANT INSTRUCTOR IN TELEGRAPHY^ S.M.E. CHATHAM*. 



During the last few months I have been observing spectroscopi- 

 cally gun-cotton flame, and have obtained results which it is thought 

 will prove of importance in calculating for the safe storage &c. of 

 that combustible. During my researches it was necessary, after 

 heating gun-cotton previously to ignition to various temperatures, 

 that I should ascertain the true temperatures of the resulting flames. 

 These I endeavoured to arrive at by subjecting platinum wire to the 

 different heats, and passing a voltaic current through it, and noting 

 the electrical resistances of the wire cold and when heated. In order 

 to find the degrees of heat, I made a series of careful experiments 

 with the same wire for a large number of known temperatures. 

 The resistances thus obtained I found did not agree accurately with 

 the formula given by Dr. Siemens. The formula that agrees correctly 

 with my results is 



This formula was arrived at theoretically by a consideration of the 

 expansion of the metal longitudinally and in diameter, together with 

 a probable value of the retardation of the current due to heat. 



Every precaution against error was adopted in the experiments. 

 The wire had a preliminary heating to a high temperature, and the 

 value of any thermo-electric current taken into account. 



The results of the spectroscopic analysis of the flame I 2:)ropose to 

 communicate in a future Number. 



* Communicated by the Author. t Practically 75 may b2 neglected. 



