in Pendulums for Gravity Measurements. 



149 



equally good. Imperfection of substance, as to strength, con- 

 tinuity, etc., is the principal difficulty and this is met by mak- 

 ing the angle of the faces as great as possible. But the errors 

 arising from the impossibility of grinding the faces to true 

 planes are reduced to a minimum by making the angle as 

 small as possible and hence there is indicated an angle neither 

 very large nor very small which is better than others. In 

 other words, if the angle is too small the edge will be entirely 

 crushed and thus broadened and made imperfect. If the faces 

 are not perfect planes it is clear that the larger the angle the 

 wider will be the limits of the horizontal projection of the 

 line of their intersection and this will be equivalent to widen- 

 ing the edge upon which the pendulum swings. 



It is interesting that this seems to be verified in practice, as 

 is shown in the results given below. Edges ground to five 

 different angles, varying from 90° to 160° were used. The 

 results show that while there is no great difference in the 

 results with the range of angles examined, it is tolerably cer- 

 tain that the angle should not (with this material — steel) be 

 less than 110° nor more than 140°. An angle of 130° will 

 probably be found the most satisfactory, combining the neces- 

 sary sharpness (when properly ground) with strength to resist 

 accidental injury or excessive elastic compression. 





Steel Knife-Edge. 





Width. 



Period. 



Gai 



P 







1 



•5008880 







2 



8839 



41 



5 



8756 



124 



6-5 



8503 



377 



10 



7626 



1254 



Angle. 



90° 

 110° 

 130° 

 150° 

 160° 





Period. 



5008875 

 8880 

 8874 



8884 

 8852 



