362 



MESSRS. R. T. GLAZEBROOK AND T. C. FITZPATRTCK 



The mean length of the threads and the corresponding values of p. are given in 

 Table IV. 



The agreement in the two values is sufficiently satisfactory. 



TABLE IV. 



To find the average cross section of the tube a thread of mercury almost filling the 

 tube was used. In nearly all cases this was the same thread as was used in finding 

 the resistance. The extreme length (/) of the thread between its curved ends was 

 found with the reading microscopes and the standard scale, while the curvature of the 

 ends was found by reading the distance between the end of the meniscus and the 

 point in which the mercury touched the glass. For a given tube this distance did 

 not vary very greatly in most of the observations, and, with one exception, VI. (2), 

 the mean value has been taken in calculating the correction 81. The method of 

 determining this correction has already been given. The temperature was observed 

 by means of a thermometer laid alongside the tube, which was generally left in 

 position for some hours before observations commenced. The microscopes and ther- 

 mometer were then read at intervals of about 15 minutes, and, when two or more 

 consecutive values for both were found to be the same, it was supposed that the 

 temperature of the mercury was that given by the thermometer. This was verified 

 in several instances by taking the temperature of the mercury after it had been 

 allowed to run from the tube into the small crucible in which it was weighed. 



As has already been stated, the mercury thread did not entirely fill the tube, and, 

 in consequence, a small correction was needed. The amount of this correction is 



; 1 ), s being the mean cross section, and ^ the mean of the cross sections 



at the ends. 



To determine the ratio s/5 1 a point was found on the tube from the calibration 

 experiments at which the actual cross section was equal to the mean value. A short 

 thread of mercury, some 3 to 4 mm. in length, was introduced, and its length 

 measured when its middle point was at the point of mean cross section : let this 

 length be I. The thread was then moved to one end of the tube and its length 



