WAVE LENCiTHS IN THE INVISIBLE IMMSMATIO SPEOTIUJM. 



155 



as the data. This I say with special rofereiice to the hirj;'i' orijiiiiai rhiiits* wliicli hii\e l>e('n drawn 

 by Mr. J. E. Keelkr, of this observatory, and which seem fo nie favorable specniiieiis of the 

 accuracy obtaiuable by this method. 



We are now prepared to test the accuracy of the various furmuUc /■oinicctiiig refraction with wave- 

 length, thougli it will be conveuieut to first prepare a table sliowinj; what this relation is in the 

 visible part of the s|)eetriiiii of the i)risin employed. 



In the following- table the deviations in the visible speutriini were measured by tlie spectrom- 

 eter, reading to 1(1" of arc, which has been already described, in which for this special pur|)ose 

 tlie bolometer was replaced by an aclironiatic observing telescoi)e with a micrometer eye- piece, and 

 the indices of refraction were computed by the usual formula. () in the ultra-violet was measured 

 by the aid of a Souet tluoresceut eye-piece, and its wave-lengtli is from Cornu. The other wave- 

 lengths are taken from Angstrom, but the unit is here the micion = .()()l millimeter := (10,000 

 times the unit of Angstrom's scale). A is here the symbol for the wave-length. 



The following indices in the visible spectrum, on wliich the computations for testing the for- 

 mulae are founded, are trustworthy to the fourth decimal |)lace here given. 



Table III. — Observed indices in visible spectrum <>J Hilf/er prism. 



Line A Ou. 76009 



C On. 6o(il8 



D, I 0//.5e8U0 



l>^ I 0//.51667 



F \ 0//. 4>-606 



H, ' 0//.3%79 



O I 0/i. 34400 



4{i^ 4ii' n.^i'' 

 47^ l.V 4.'>" 

 47° 45' 15" 

 4ft°2r05" 

 4f.o 44' 15" 



.'•iO° 34' 0.=i" 

 52° 43' 00" 



1..5714 

 1..5757 

 1.5798 

 1. 5P62 

 1. 5899 

 1. fi070 

 1. 6266 



A sinooth curve drawn through points whose positions are given by the above table, repre- 

 sents with accuracy the relation between n and /\ in the visible i)art of the spectrum. This method 

 is, however, obviously inapplicable to the very extended invisible portion below the A line, and 

 accordingly attemi)ts were first made to effect the determination of corresponding indices and 

 wave-lengths by extending the curve derived from the abo\e observations by means of formula;. 

 Several fonnuliB have, it will be remembered, been proposed by i)hysicists, expressing u as a func- 

 tion of A, and containing constants which are to be determined by observation, but it has never 

 hitherto been possible to test these formulae far from the visible spectrumr, whence their constants 

 have been in fact derived. This desirable test we are now- prepared to apply. 



The simplest as well as the most widely used formula is that of Caticht, which, as it is com- 

 monly written. 



(^n = a + l,+ l,j 



contains three unknown quantities, requiring for their determination three simultaneous equations. 

 Selecting the lines A, D, and H for this purpose, we have from the table just given, the three 

 equations, 



1.5714 = a -f 



(0.76009)^ "*" (0.7000!))* 



•These original charts were exhibited to the members of the National Academy of Sciences, at Washington, in 

 April, 1883. The engraving here given in illustration being on a mncli reduced scale, will merely indicate the exact- 

 ness of interpolation possible by the originals. 



