Increased Accuracy in the Use of Bifilar Suspensions. 443 



to rest within twenty seconds. In order to eliminate these 

 errors, it was necessary to lake the mean of a number 

 of observations on any creep under the same repeated 

 conditions. 



Below are the series of observations o£ the creep for 

 various values of T. In Table I. the full set of original 

 observations are given in order to illustrate the method of 

 procedure in calculating the mean values. For other values 

 of T the results were obtained in exactly the same way, and 

 merely the mean sets are shown. 



Between the determination of each of the columns, the 

 suspension was allowed to rest, and then the deflexion was 

 again produced and reversed every minute a large number 

 of times, in order to eliminate the effects of the previous 

 seven or eight minutes' distortion. 



To determine the constants h and D it is convenient to 

 calculate the successive differences in each case, and then take 

 the average of these. Let ^,. mean the dynamometer reading 

 when t = r minutes after the last reversal, and we get the 

 folloAving average values for the creep when T = l from 

 Table 1.:— 



#2 — a?x = "009 cm. 



a, -^ = -008 „ 



A' 4 — J' 3 = *005 „ 



x-o— a** = '005 „ 

 * e -* 5 =-004 „ 



For the various values of T we get in exactly the same 

 way all the results in Table IT. In the cases of T = 2 and 

 T = 7, only a few sets were taken. At the foot of each 

 column is a figure representing the relative weight of each 

 series. 



Table II. 



Observations for 



T = l/2. 



T = l. 



T = 2. 



T=5. 



T=7. 



T=9. 



X — X 



•007 

 •006 

 •004 

 •004 

 •003 



•009 

 •008 

 •005 

 •005 

 •004 



•011 

 •007 

 •008 



j -010* 



(1) 



•013 

 •011 

 •003 



~(2) 



•014 

 •011 

 •005 

 -034 

 •007 



(1) 



•014 

 •010 

 •007 



•oos 



•005 

 •004 

 •004 



(3) 



x z~ x i 



X^—X% 



X- — x< 





x~ — x c ., 







Weight 



(2) 



(3) 



* This value is for x 6 — x v 



