500 Prof. Q. Majorana: Theoretical and 



As abscissae I take the successive intervals o£ time Ci S 2 , 

 S2 C 3? C 3 S 4 , S 4 C 5 , . . . . , all equal, elapsing between the 

 single observations with mercury and witliout mercury alter- 

 nately. As ordinates I take the scale's positions of rest, 

 determined each of them with three readings of oscillations. 

 I join the points obtained in this way with two lines. 

 These have a descending direction caused by the gradual 

 displacement of the rest position of the balance, which dis- 

 placement is a consequence of slight variations in the 

 temperature. But the with-mercury line is always, with its 

 points, above the without -mercury one. This means that the 

 presence of the mercury always makes the leaden ball m 

 seem lighter. In the same fig. 4 the 14 vertical segments 

 represent the various successive means obtained from the 

 represented series of observations. For briefness sake I 

 shall not report all the other diagrams corresponding to five 

 other series of observations that were made, together with 

 the one described in fig. 4, on July 20 and 21. I shall say 

 only that, taking the general mean of 57 partial means, I 

 find as the value of the displacement of the balance position 

 of rest, due to the presence of the mercury, 



mm. 0-358 + 0-012; 



the probable error 0*012 has been calculated by the method 

 of least squares. The direction of the displacement indicates 

 a diminution of weight, that is to say, absorption of the 

 terrestrial force of gravitation on the leaden ball through the 

 mercury. 



The balance sensibility in the course of the aforesaid 

 experiments was constantly equal to 171 mm. per mg. 

 Therefore that displacement corresponds to a variation of 



mg. <>-358±0-012 = mg _ Q()m ± . Qmj ^ 



Correction of the observed effect. — However, it must be 

 stated that in the experiment carried out in this way, several 

 causes intervene, and superposing themselves with their own 

 effects upon the phenomenon sought, modify the result 

 notably. I cannot in this brief exposition discuss such 

 causes in detail ; but among them I select those which have 

 sensible effect, and I construct with them the following table 

 (each bears its own sign) : — 



