VALUE OF ACCELERATION OF GRAVITY— MENDENHALL. 23 



cause of the irregularities in Ring- II, it is impossible to draw any conclusions as to the magni- 

 tude or even the sign of the collection which should he applied to the observed period, and the 

 best that can be done is to take the mean. 



To determine <j we have then the following corrected values: 



From these 



Considering the irregular behavior of King II these values are in very satisfactory agreement. 



The results of this work can be taken, it is hoped, both as throwing light on the reliability 

 and practicability of the method and as furnishing a new absolute determination of g, though in 

 order to compare the value here found with previous results, either the ring pendulum must be 

 swung in Washington or a relative determination made between there and Madison. It is 

 believed, however, that the work demonstrates: 



(1) The possibility of making glass-hard rings of steel of sufficiently accurate figure for 

 the purpose. 



(2) That some rings show a very satisfactory uniformity of period for different positions. 



(3) That even where considerable asymmetry exists the mean period corresponds very 

 closely to the period of the corresponding perfect ring. 



(4) That different rings will yield extremely concordent results. 



(5) Finally, that with rings prepared with the degree of precision which has been attained, 

 the method will be valuable as a check upon determinations of gravity made by methods 

 hitherto in use. 



In conclusion, I wish to acknowledge my indebtedness to the University of Wisconsin for 

 the facilities accorded for the work, and to Mr. H. G. Fisher, mechanician of the department 

 of physics, for faithful and painstaking assistance in the construction of most of the apparatus. 



August, 1904. 



