the Motion of Spheres through Liquids. 70.") 



These results are shown graphically in fig. 1, the upper 

 curve showing ^ and the lower /x-l- It is seen that /jll 

 remains approximately constant up to the radius r^ = "04? *»•« - 



cm 



Fiff. 1. 



2.0 

 1.8 

 1-6 

 14 

 '•2 

 1-0 

 ^0-8 



1 1 

 Rose metal in Colza oil 











jl by Stokes law - upper cu? 



jo by £aden6er<? - lower curt 

 ~ jcby Foiseuille. - 0-696 



Diameter of tube - s*7 cm 

 — Length, of fall - 53-4- 



Critical radius -oee - 



Temperature & U ° 



■re 

 T. 









,. 





















































} p o 



0-4 

 0-2 

























































r. 



•03 



•04- -05 



Radius in CM. 



•06 cm. -07 



The average of the values for radii less than r l gives 0*6930, 

 agreeing closely with the value 0'6957 given by the capillary 

 tube method. This is an excellent confirmation of the validity 

 of Ladenberg's correction for an oil of less than 1/2000 the 

 viscosity of Venice turpentine. 



As it was conceivable that the rise of the lower curve 

 might be due to the failure of Ladenberg's correction rather 

 than to the influence of the inertia terms directly, a further 

 test of the validity of the correction was made with tubes A 

 and C. The results are plotted in figs. 2 & 3 (p. 764), where 

 the curves have the same significance a^ in the previous figure, 

 A and A', G and C, referring respectively to the values of 

 /as and of /zl obtained with the two tubes. The curve A' 

 begins to rise at about ?'=0'32 cm., that is, when the ratio 

 r/R is about 1/10. Since the tube is very small this rise is 

 probably due to the failure of the correcting factor, hence it 

 is not effective for values of r which make the ratio greater 

 than this. With tube the ratio becomes 1/J only when 

 r = 0'16, hence the rise of C at r = '045 can be attributed 

 only to the inertia terms. This, it will be noted, is very 

 nearly the value of r x obtained with tube B. It would appear 

 from this, since the computed value of r is *066 cm., that the 



