Observations on various Liquids, 



235 



same time, one giving an observed lower limit of potential, 

 and the other an observed higher limit : these are determined 

 in the same way as the preceding, the index being clearly 

 beyond the sighted position in the one case at the instant of 

 the signal, and clearly short of it in the other case, always with 

 rising potential. 



Although the details of this method were adopted as the 

 best that occurred to me in the course of a good deal of pre- 

 liminary work, the final observations were not quite satisfac- 

 tory. The effect in the polariscope varied in many cases so 

 slowly that the observed limits of potential were too far apart. 

 The compensator itself was subject to small irregularities, which 

 it was almost impossible to prevent or to neutralize perfectly. 

 Whether these difficulties are inseparable or not from the 

 method, and how far they might be overcome by a good 

 choice of materials and by very careful work, I am not yet 

 able to judge. 



48. Carbon Disulphide. — The following Table contains the 

 results of two early sets of observations. Columns A and B 

 give the numbers of pounds attached to the first and second 

 plates respectively of the fixed compensator ; column C gives 

 the observed range of corresponding or sensibly neutralizing 

 potential ; D and E the observed lower and higher limits of 

 potential ; Gr the potential as inferred from the preceding 

 numbers. Column K gives the observed increments of poten- 

 tial corresponding to a constant increment of birefringent 

 action (2 pounds on second plate). 



A. 



B. 



0. 



D. 



E. 



a. 



K. 



1 







57 



56 



58 



57 





2 







67 



66 



68 



67 





4 







86,87 



85 



88 



86£ 





7 







113, 115 



112 



116 



114 





11 







136, 138 



135 



139 



137 . 





14 







151, 154 



150 



155 



152| 





2 







70 



69 



71 



70 



2H 



2 



2 



91,92 



90 



93 



91 £ 





7 







112,116 



111 



117 



114 



12£ 



7 



2 



124, 128 



123 



130 



126i 





14 







154 



153 



155 



154 



9* 



14 



2 



163, 164 



162 



165 



1634 





The next Table contains an unaltered record of the most of 

 my last day's work with CS 2 . All the conditions of observa- 

 tion were kept as constant as possible. The columns are de- 

 signated exactly as in the former Table, 



