in Pendulums for Gravity Measurements. 147 



upon which swings will only rarely be made in order to detect 

 any deterioration which may take place in those in daily use. 

 Knife-edges of different material may also be used if such a 

 course is found to be desirable. 



In the reversible pendulum for absolute measures, in addi- 

 tion to the advantages already described there is also the very 

 important fact that the measurement of the length of the 

 pendulum is likely to be more accurate. Whether the knife- 

 edge is a part of the pendulum or of the fixed support, a 

 certain amount of elastic compression will always take place 

 when the weight of the pendulum is upon it and this is likely 

 to be relatively greater the more perfect the edge. If the 

 knife-edge is a part of the pendulum the length of the latter 

 must be greater when vibrating than when at rest and sup- 

 ported with the edge free for length measurement. 



The measurement of the distance between the two knife- 

 edges of a reversible pendulum as ordinarily made is a matter 

 of much difficulty. It is believed that without resorting to 

 extraordinary methods the distance between the two planes of 

 the newly constructed reversible pendulum will be capable of 

 measurement with higher accuracy and if necessary or desira- 

 ble, recourse may be had to the method of Michelson and 

 Morley for relating the distance between surfaces to the length 

 of light waves, 



Perhaps the most important gain thus far from the new 

 arrangement is that it has made it possible to investigate the 

 knife-edge, as to form and material, with an ease and thorough- 

 ness hitherto unattainable. As long as the knife-edge forms a 

 part of the pendulum it is impossible to study the effects of 

 variation in its angle, width or material because every such 

 alteration must necessarily alter the vibration-period by chang- 

 ing the mass and form. With the new form this difficulty no 

 longer exists ; the vibrating body remains constant in mass 

 and configuration and any change in its period is due to the 

 influence of the knife-edge. 



Some of the results already obtained are of sufficient interest 

 to justify their publication in advance of a full and complete 

 investigation now in progress. 



In order to investigate the effect of a wearing or slight 

 flattening of the edge, such as may, and in fact does result 

 from long and not too careful use, a steel knife-edge was used, 

 the plane, forming a part of the pendulum, being of agate. A 

 steel edge was used on account of the greater ease with which 

 it could be manipulated in grinding. The agate knife-edge 

 has been in general use, but experiment showed that both 

 steel and agate being made as perfect as possible, the vibration 

 periods were essentially the same. The angle which the two 



