F. Dellmann on Atmospheric Electricity. 411 



will be seen immediately, this hope has only partially been ful- 

 filled. The sensitiveness lost in the new arrangement was quite 

 unimportant, as it could easily be replaced by a glass thread of 

 smaller thickness. During last summer I used a glass thread 

 27 centimetres long, which received from a battery of 64 zinc- 

 copper elements a deflection of 44°. 



Torsion -measurements were again made with the new arrange- 

 ment, and the result surpassed all expectation. A few series of 

 numbers may be here given as examples. The instrument 

 was charged with an arbitrary quantity of electricity, and then, 

 in all cases with the same charge, the torsions at 80, 70, 60, 50, 

 40, 30, 20, 10, 20, 30, 40, 50, 60, 70, and 80 degrees were 

 measured. 



These two series were made under circumstances which might 

 give rise to differences in reading off small angles, the first series 

 being made by daylight, the other by artificial light at night. 

 Two other series made with another instrument may also be given 

 here, since in the second the strip was shorter and narrower than 

 before. 



Although the agreement between the series made with one 

 and the same instrument was satisfactory, the non-agreement of 

 the series made with different instruments called for closer in- 

 vestigation. It was this investigation which first led to a com- 

 plete knowledge of the instrument. By repeated torsion-mea- 

 surements under variation of the principal components, the strip 

 and balance-arm, with respect to their dimensions and distance 

 asunder, it was found that for the gradation of the scries the di- 



