22 



Mr. G. J. Burch. 



[Nov. 21, 



brated by the old method, but bad become " sticky " at its best part, 

 and the new process was therefore applied to the portion below the 

 obstructions and extending beyond the region of equal sensitiveness. 



TV 4- 



JJistance 

 from the 

 reierence- 



circle, 

 in rams. 



Upward 

 excursion 

 -N. 



Subnormals. 



Downward 

 excursion 





Value per c 

 of the sub 



Relative 



AT j AT' 

 JSp + JN p 



1 



entimetre 

 normal. 



In volts. 





Tube sticky. 



Value per cm., as found 











by previous calibration, constant 



} " 



-01333 





throughout this part. 









20 



-13-0 



+ 42 -0 



55-0 



1 -ooo 



-01333 



25 



-16 -5 



+ 38'5 



55 -0 



1-000 



-01333 



30 



-21-5 



+ 33*0 



54 -5 



1-009 



'01345 



35 



-25 -0 



+ 29-0 



54-0 



1-018 



-01357 



40 



-30 5 



+ 215 



52 -0 



1-058 



0-01411 



45 



-34 -5 



+ 16-0 



50 -0 



1-100 



-01466 



50 



-37 -0 



+ 9-5 



46 -5 



1 -183 



-01577 



The fourth column shows the initial velocity of the excursion 

 caused by the same difference of potential, for seven equidistant 

 positions of the meniscus. But the exact value of this difference of 

 potential is unknown, because the zero-positions of the excursions 

 are not recorded. A third photograph is therefore taken in order to 

 determine the value in volts of a centimetre of the subnormal at some 

 one of the positions in the table. To do this, the potentiometer is 

 set accurately to give some known P.D. The zero point of the 

 meniscus is brought about 30 mm. below the reference circle, and the 

 excursion is directed upwards. The subnormal to the resulting 

 curve is measured at a point 25 mm. from the reference-circle, i.e., in 

 the middle of the available space. It would be difficult to make the 

 excursion start exactly from the 25 mm. mark, and besides it is 

 easier to place the tangent-line accurately against the curve at a 

 point some distance from its origin * The value of the acting P.D. at 

 the point measured is found by subtracting the P.D. corresponding 

 to the distance through which the meniscus has risen, from the P.D. 

 indicated by the potentiometer. 



In the case quoted, the acting P.D., at a point 25 mm. from the 

 reference circle, was 0"02 volt, and the subnormal to the curve at 

 this point was 15 cm. 



* Note added November 19, 1895. — The shape of this curve at its origin show s 

 whether the instrument is dead-beat or not. If the velocity of the meniscus 

 increases after the excursion has begun, the capillary should be rejected. — G-. J. B. 



