26 



Prof. Guthrie on Bubbles. 



TABLE a. 



From glass, air-bubbles through water. 

 T=23C. 

 B = 767millims. 



[Jan. 19, 



It would at first seem as though there were a well-marked difference 

 depending upon the value of gt. But in this method of experimenting 

 there is a possible maximum error of two bubbles in each case, or an error 

 of four bubbles in the comparison of two instances. This nearly covers the 

 observed discrepancy. To set this point at rest, experiments were made 

 with a larger number of bubbles as follows. The vessel taken for M was a 

 100 cub. centims. flask. The water in A was each time filled up to a. 

 Only those two of the values of gt which gave the most widely differing 

 results in the preceding Table were reexamined, viz. gt= 0'33 and gt= 1 -00. 

 A thread was fastened to the end of the siphon D, so as to deliver its con- 

 tents in a series of very rapid and minute drops. 



