GRAVITY MEASUREMENTS. 
41 
in Europe, as will be seen by examining the accompanying 
Table A of values reduced to Washington, the results rang¬ 
ing from 980.047 to 980.285, with a mean of 980.107 (omit¬ 
ting Kater’s discordant result), not very different from the 
adopted provisional value. Of course these results involve 
the errors of the relative connections, which in some cases 
are not satisfactory, but these are thought to be small in 
comparison with the discrepancies in the absolute measure¬ 
ments themselves. 
Table B gives for each station the values of g derived from 
each pendulum, with the mean of the three and the differ¬ 
ences from the mean. The largest single discrepancy of 
any result from the mean is seen to be .006 dyne, or about its 
Te oVo o part. 
