﻿Measurements of Pendent Drops, 

 Table II. 



427 



* 



x\ 



x% 



bx 2 . 



ex 4 . 



•• 



y'—y- 



12-088 



27-731 



1-4240 



•2710 



•0020 



•2690 



-•0015 



5-169 



8-934 



•6023 



•2041 



•0011 



•2030 



+ •0015 



1-921 



2-387 



•2217 



•1468 



•0006 



•1462 



+ •0027 



•587 



•491 



•0690 



•0989 



•0003 



•0986 



+•0001 



•134 



•068 



•0158 



•0604 



•0001 



•0603 



-•0002 



•019 



•005 



•0022 



•0313 



•00003 



•0313 



+ •0008 



•001 





•000 L 



•0117 

 •0016 

 •0008 





•0117 

 •0016 

 •0008 



-•0018 

 -•0029 

 -•0017 



•001 







•0121 





•0121 



+ •0006 



■013 



•003 



•obi 6 



•0278 



•00002 



•0278 



-•0007 



•103 



•018 



•0124 



•0554 



•0001 



•0553 



-•0012 



•480 



•376 



•0567 



•0925 



■0002 



•0923 



-•0002 



1 -(526 



1913 



•1957 



•1389 



•0005 



•1384 



-•0031 



4-492 



7415 



•5270 



•1948 



•0010 



•1938 



+ •0003 



10706 



23-600 



1-2510 



•2602 



•0018 , 



•2584 



+ •0009 



From the above table we have 



^ = 37-339, 2^=72-972, 2^=4*3796. 



To determine b and c, the principle of least squares gives 



2(y — bar — c#*) 2 = min., 



the condition for which, viz. : 



gives as the normal equations for b and c 



btar-\-ctar = t^y 

 bXa* + cZrf = %x A y. 



Substituting the numerical values, and solving, 



b = -HSl, c = — "000375, 



giving as the equation to the meridian curve over the 

 portion considered 



y = -ll£l.i- 2 --000375.r 4 . 



Using the same values of ,v as in Table I. the R.H.S. of 

 this equation was evaluated as shown in Table II., the last 

 column of which shows, as before, the difference between 

 the observed and computed values of //. It will be noticed 

 that, whilst the deviations are of the same order as those 

 corresponding to the parabolic formula, they are much more 



