58 PHILOSOPHICAL SOCIETY OF WASHINGTON. 
ponding epochs by ¢,/ and,” the rate of the time-piece will be 
given with sufficient accuracy for interpolation by the equation 
At” — at’ 
Mr. Paut thought it an objection to this method that in arrang- 
ing the groups valuable stars might be lost, so that in a limited 
time the accuracy of the results would be impaired by the smaller 
number of observations ; moreover, the method did not furnish the 
computer with a clear idea of the performance of the instrument. 
Mr. Hatz said that he liked the method, and that he thought it 
especially good for time work. He had discovered the method once 
himself, and he knew that it had also been used by Prof. Ormond 
Stone. 
Mr. KumMeE t said that in connection with this subject he had 
investigated the question of the advisability of using stars towards 
the pole for time determinations; that is, he had examined the 
weight co-efficient formule to see at what distance north of the 
zenith a maximum value would be obtained. He found that in 
general the limit of declination was about 60°. 
Mr. Paut thought that every weight-formula should take account 
of the increase of atmospheric disturbance with increase of zenith 
distance. 
T= 
Mr. KumMe t then read the following paper entitled 
CAN THE ATTRACTION OF A FINITE MASS BE INFINITE? 
In Price’s Calculus, vol. III, art. 201, discussing the result for 
the attraction of a thin rectangular plate on a particle external to 
it and in its own plane it is found that if the attracted particle is at 
an angle of the rectangle the attraction is infinite. Price’s method 
of determining the attraction of plates consists in integrating be- 
tween the proper limits the following differentials : 
ax = Se qd ) 
(a? + y')° ‘ 
a’ Y = mor es Es 3 
(a? ue y’)” y 
where X = x — axial component of attraction; 
Y = y — axial component of attraction; 
= mass of attracted particle ; 
6 = density of attracting mass, supposed homogeneous ; 
t = thickness of the plate. 
