204 



In oil of lavender and of sandal-wood there was some indistinct- 

 ness in the line H which renders its index a little uncertain. It may 

 be owing to this circumstance that the assumption of that index may 

 have occasioned the discrepancy between theory and observation. 



In oil of spikenard the accordance is good. In benzole the discre- 

 pancies are too great. 



Bisulphide of Carbon. Dale and Gladstone. 



In the first of these media the differences are greater than can be 

 fairly allowed to errors of observation. 



In the second case it is yet more clearly apparent that the theory 

 is defective. The ray C was not observed ; but the theoretical index 

 is evidently in error to a large amount, as it is even lower than that 

 of B. The indices for D and C are perhaps within the limits of 

 error : but that of E is too much in defect to be allowed. 



