100 G. H. KNIBBS. 
in which therefore c’ =c/(ug), we obtain 
4 
m=c B gee: (9) 
If the density of the liquid in the manometers (2) be identical 
with that of the effluent liquid (p), /p will be unity. ; 
By formula (8a) and (9), I have ascertained where possible, the 
values of m from Poiseuille’s measurements with tubes A tof — 
Those given hereunder are derived from a combination of numerical — 
and graphic methods. Through the points whose codrdinates 1/7 
and P 7' were drived from Poiseuille’s series, as for example, the ” 
series A,, A;, B;, C;, shewn in Fig. 2, a mean line was drawn. — 
Two members of the series most nearly agreeing with this line 
were usually then selected and the values of ¢ and C found by the 
formule 
Se Fa kale a0 
~~ 
C=P,T, - 7 (11) 
When for each tube designated by the one letter as Ay, 4s, 
c — been determined, its multiplication by the fractional fact 
in (9)—a constant for the tubes so designated—gave the values of i 
m. The radius used was not the mean for the whole tube but tht 
mean radius’ at it entrance. ' 
Values of m deduced from Poiseuille’s experiments. 
Length i Leng 
abe, Mengt Madina) Valnon ing, Length Satan | 
A, 255 00708 104 3B, ‘39 00567 i 
yes S 5 102 C, 60 00427 Pera 
Ppa e 115 F, 20:00 -03267 : 
ie eee 
Ay AO Tk 
LR Ol ee Fig ke 8 re 
y, Pgeaee F, 0 
Mean 1-14 : ad rejecting C.P.,P. 113 : 
eager t 
1 Actually h,c' and C’ were used. Thus in (10) c’ replaced c, and hyP : 
In (11) C', hh and ¢' replaced C, P and c. . 
greatest and least radii. If the section be truly an ellipse the ™ 
radius so defined will be that of a circle of equal area. ; 
