?9' 



8 



II 



/l 



^<yo 



( ' Ö3 ) 



The aiiiioxed figure represents 

 uraphicallv u part of these read- 

 ings for both sets. Tlie irregular 

 (lilferenees, ma}' apparently be 

 ascribed to errors of obser\ation, 

 which near the limit of absorp- 

 tion will be somewhat larger 

 than at other places, owing to 

 the smaller intensity of the light. 

 They do not amount to much 

 more than 1 [i\i. A deviation 

 of the kind which we might 

 expect from the source of errors 

 supposed by Bates, would reveal 

 itself, near the limit of absorp- 

 tion, in a displacement of the 

 band towards this limit. Such 

 a displacement is not at all 

 indicated by these obser\ations. 

 Let us consider what apparent 

 displacement must ha^e taken 

 for the anomalous rotations which are found in 

 This mav l)e found by supposing for a moment 



ioo 



510 



sorit's with water 



solution 



limit (if absori>tiou 



}»lacc, to account 

 the measurements, 

 that the rotati<m of the salt is normal, and by putting it equal to 

 that of water. If for instance we start from the value v^,;, =7.1 for 

 A, = Goto ') and we call \ the wave-length, where the l)and ought 

 to have appeared with the solution, if it appeared with \vater at 

 ;..^ = 519, then we tind by a simple calculation ;.i = 509, while we 

 ha\e obser\'ed l^ = 5Ö0. According to what has been said before a 

 dis])lacement of the band of 9 up cannot be apparent. Hence the 

 validity of the results obtained before is not alFected by the error 

 supposed by Bates. 



1) Comm. N'J. 70 p. 4; Froc. Royal Acad. 1901/0:2 p. 3iO. 



