779 
curve bbb constructed on the red side of the band shows at once 
-that, in this case, 
À 02 = 6,36.10~ 5 cm. (10) 
If we attempt however to carry the calculation any further, wé 
are soon compelled to admit that with the data at our disposal it 
is useless to try to attain reliable results. Hence, hitherto we have 
dealt only with the method to be adopted in proceeding to a second 
approximation. To perform the complete calculation on the lines 
here indicated would require numerous experimental data, of a very 
high degree of accuracy. 
B. Similar results are obtained when we consider the behaviour 
of the other solutions investigated by M. Cotton; for instance that 
of a mixture of chromium alum and sodium tartarate solution 
(Cotton, l. c., pages 415 and 416). It is unnecessary to consider 
this example in detail. The agreement of experimental evidence with 
«mono-electronic» anticipations is even less than in the case con¬ 
sidered above. 
§ 12. We now proceed to consider the ellipticity produced by 
an absorbing active solution. Let us see to what extent the varia¬ 
tion with the wave-length of the angle cp as observed by M. Cotton 
can he correctly represented by equation (4) of Art. 8. To ascertain 
this in the case of the double tartarate of chromium and potassium 
solution, already dealt with in § 11 A, the value of the expression 
U — L 2 ) 2 4-r 2 ^ 2 . 
---•agio tg(45° — <p) 
( 1 ) 
has been computed from the values given on p. 407 of M. Cotton’s 
paper. Writing R for the expression (1) we get from equation (4) 
in Art. 8. 
R — 21 D I T log 10 e = Const. (2) 
The results obtained run as follows: 
l (p R 
5,22 . IO“ 5 cm 1° 25' 1,84.10~ 16 cm^ 
5,40 . 2° 33' * 2,38 . 
5,81 4° 54' 4,50 . 
5,89 . 3° 40' 3,85 . 
* Read from the curve. 
