192 W. F. Magie — Method of Measuring 



Absolute Alcohol, sp, 



, gr 



. 0-7934 at 1. 



5-6° 



c. 



k 

 3-437 



l 

 45-5 





a- 

 5 760 





Temp. 

 16-6° C. 



3-45 



47-5 





5-808 





14-4 



3-442 



50 





5-789 







3-407 



60 

 Mean, 





5-698 





12-8 





5-764 







Olive Oil, sp. 



gr. 



0-914 at 18° 



C. 





k 

 3-917 



1 

 55 





a 2 

 7-491 





Temp. 

 18-6° C. 



3-872 



52-5 





7-317 





19 



3-914 



47-5 





7.447 





18 



3-892 



52-5 

 Mean, 





7-387 

 7-410 





18 





Petroleum, sp. 



gr. 



, 0-808 at 16 c 



'C. 





k 

 3-682 



1 

 60-5 





a 2 

 6-641 





Temp. 

 15-4° C. 



3-753 



57-5 





6-891 





15-4 



3-717 



52-5 





6-744 





15.2 



3-717 



52 





6-744 





16'5 



Mean, 6-755 



Each of the above values of h is the mean of several (usually 

 four) independent measurements. The figure in the third deci- 

 mal place is the result of the division by which the mean was 

 obtained. The original readings were made by estimation to 

 hundredths of a millimeter. 



In order to compare the above results with those given by 

 other observers, it is necessary to reduce them to a common 

 temperature. For this purpose I have used the coefficients 

 given by Brunner or Frankenheim and Sondhaus. The final 

 mean for a 2 obtained for water at 0° is 15-54. Laplace uses 

 15-36, which is the value usually adopted by the older obser- 

 vers. In a recent series of measurements by a method not 

 involving the contact angle,* I obtained 15'00 as a value for a 2 . 

 Mr. Durell, Fellow in Experimental Science at Princeton, who 

 has kindly undertaken for me a series of measurements of a 2 

 from the rise of liquids in capillary tubes, has obtained the 

 value 15'62. The discrepancies in these different results, while 

 they show that much remains to be done in perfecting the vari- 

 ous experimental methods, still show nothing as to the validity 

 of the assumption that the contact angle is zero ; for, if this 

 assumption be untrue, the value of a 2 obtained by methods not 



*Wied. Ann., xxv, 421. 



