180 Dr. W. F. Magie on the 



tion is supported by the observations of March 10 and April 24. 

 The measurements were then made while the film of liquid 

 still covered the lens' surface, so that the supposed irregula- 

 rities at the circle of contact were far less likely to occur. 

 The settings were more difficult and the readings not quite so 

 accordant, but the agreement of the results by the two methods 

 is such as to indicate that the cause of the discrepancy in the 

 results of the other days was not operative. 



This view obtains further support from the results of another 

 series of measurements on ethyl-alcohol*. By the " q " method 

 the value then obtained for a 2 was 5*576 at 23°. By the " k" 

 method, in which the bubbles were formed under a flat plate, 

 the values were 5*803 and 5*786 at 22°. From the ordinary 

 method by the use of capillary tubes the value was 5*805 ; 

 and by a method in which the radius of curvature of the 

 meniscus in a capillary tube was measured directly, and which 

 did not involve the contact-angle f, the value was 5*859. In 

 the last two methods the disturbing effect of the water in the 

 surface-film would be felt as well as in the " k " method. 

 Hence arises, I think, the substantial agreement of the results 

 by the last three methods, and also the difference between them 

 and the results by the " q " method. 



The results of March 10 offer no evidence of a finite contact- 

 angle ; and though those of April 24 differ by more than one 

 per cent., yet this difference is mostly due to the values given 

 by the last bubble. If these be omitted, the means from the 

 four others are 5*592 and 5*560. The difference is too slight 

 to demonstrate the existence of a finite contact-angle. 



Other Liquids. — In the cases of chloroform, methyl-alcohol, 

 formic acid, and benzine, the agreement of the results is such 

 as not to indicate a finite contact-angle. The variations of 

 the values of a 2 in the case of formic acid, and the way in 

 which those variations were shown by both methods of obser- 

 vation, are worthy of notice. 



In the cases of acetic acid, turpentine, petroleum, and ether, 

 the results indicate the existence of a finite contact-angle. A 

 rough calculation of the value of that angle was made in each 

 case by taking the value of k which would give the mean value 

 of a 2 given by the " k " method, and using it with the mean 

 value of a 2 given by the " q " method to calculate the value 

 of co from formula II a. In the numerical values thus ob- 

 tained little confidence can be placed. They are only given 

 as rough approximations. The results are collected in the 

 following table : — 



* Magie, Proc. Am. Ass. New York, 1887, p. 79. 

 t Magie, Wied. Ann. xxv. p. 421 (1885). 



