98 Pi'of. J. R. Young on the Combination of 



If the blue radiation or white light liberates iodine or bro- 

 mine, these elements would evaporate or combine with the 

 silver surface immediately beneath. If we take the first idea, 

 how comes it that the red radiation re-establishes the com- 

 pound in its primitive proportions ; and, in the second case, 

 how does it happen that these rays are capable of decomposing 

 the surface beneath, liberating the iodine or bromine, and then 

 combining them again with the upper surface? It is impos- 

 sible to admit that the red radiation is endowed at the same 

 time with the properly of separating and the property of re- 

 uniting the same elements. We must then attribute it to a 

 particular force — electricity perhaps, which might accompany 

 each radiation, and which, under the influence of the one, 

 would act positively, and negatively under the other, without 

 changing the chemical compound. In one case this influence 

 would give the affinity for mercury, and in the other destroy it. 

 At all events, we must look for another explanation of the 

 phsenomenon than the one which has hitherto been received, 

 viz. the decomposition of the iodide of silver by the action of 

 light. It is true that light decomposes iodide of silver, form- 

 ing a subiodide, but this seems to require a longer time than 

 that during which the surface is endowed with the property 

 of attracting the vapours of mercury. In fact, the last pro- 

 perty is communicated nearly instantaneously, which is not the 

 case for the decomposition of the iodide by the action of light. 



XV. On the Combination of the Theorems o/'Maclaurin and 

 Taylor. By J. R. Young, Professor of Mathematics in 

 Belfast College^. 



IN the application of Maclaurin's theorem to the develop- 

 ment of a function F {a-\-x) in a series proceeding accord- 

 ino- to the powers of .r, we are always directed to differentiate 

 the function F in reference to x, and then to put zero for x in 

 the several results. There is an unnecessary expenditure of 

 .symbolical work, and therefore of time in following this di- 

 rection ; for it is an axiomatic principle, which I have else- 

 where announced t» that whether we differentiate 'F{a + x) in 

 reference to ,r, and afterwards make x zero, or make x zero 

 first, and then differentiate in reference to a, the results are 

 identical; that is, 



r rf"F(fl + .r) 1 . t?"F(a) 

 dx'* J da'' ' 



• Communicated by the Author. 



t Mechanics' Magazine, Jan. 15, 1848. 



