240 



sured from bottom to bottom, a correction must be app- 

 lied to the measure. Supposing the meniscus half-spherical 

 this correction will be for both ends 



when r signifies the radius of the tube. 



The length of the tube 10,3 cm, the diam. 0,5o. 

 See Table of Observations Series I. 



If we put the 3 first results in a curve, in which the 

 abcissa signifies the distance of the point from the meni- 

 scus (2,5 mm = 1 cm) and the ordinata the quantity of 

 water drawn up in mg, (0,8 mg = 1 mm), we shall re- 

 ceive the curve, which Pl. I is marked a b c and for the 

 intensity the curve cl e f. The zero point has here ad- 

 vanced 2 cm forwards, with the same value of the abscissa 

 as abové, whereas 1 mm of the ordinata signifies 8 sparks. 

 From the two latter observations we shall receive in the 

 same way the curves a' h' and cl' f. 



The quantity drawn up seems to be nearly proportio- 

 nal to the intensity. It diminishes speedily with the in- 

 creasing distance. It is probably inverse proportional to the 

 square of the distance. 



If the observations are calculated after the formula: 



where m signifies the quantity of water and s the number 

 of sparks, r the distance and h a constant and if the method 

 of the last square is applied Te receives the value of 



31,21. 



But if we exclude an observation from the 2'^° results, 

 it will become 2,47 instead of 2,98, and Te shall be 



Ä; = 42,3 8 

 which will give the observed quantities with a probable 

 error of O, o 8 mg. 



