356 J. H. Long — Polarization of Tartrate Solutions. 



In the other cases the observations lead to a curve as ex- 

 pressed by 



[a] = (a) + Ag + Bg\ 



in which the letters have the same meaning as before. 



From the first three solutions containing potassium, nitrate I 

 have calculated according to this 



[a] = 22-°20 + 0-048^+0-0008/. 



I have not thought it necessary to find an interpolation formula 

 for all the solutions. It is evident from an inspection of the 

 curves that they intersect the base line at g = only in a few 

 instances, and the deviations are usually greater than could be 

 referred to errors of observation. 



As mentioned above, the addition of a potassium salt in- 

 creases the rotation while sodium salts diminish it. The num- 

 bers as given seem to show no relation beyond this. However, 

 if instead of giving the rotation for solutions containing equal 

 weights of active and inactive substance, they be calculated for 

 equal numbers of molecules in solution the results obtained 

 show certain regularities which must be more than accidental. 

 These desired values can be obtained readily and accurately by 

 graphic interpolation and in the following table they were so 

 derived. 



