Stokes's Law for the Motion oj Spheres in Liquids. 755 



The resemblance of Saturn's rings to the Zodiacal Light 

 is briefly indicated by Kant in a short chapter of his ' Theory 

 of the Heavens,' in which he accounts for its origin by- 

 assuming that the fire of the sun raises from its surface 

 vapours similar to those which formed Saturn's ring, and 

 by their motion around the sun formed an expanded plain 

 in the plane of the sun's equator, or in the figure of a convex 

 lens. 



Modern investigators have since carefully observed this 

 singularly interesting object, and mostly agree that it is a vast 

 accretion of cometary and meteoric particles from outer 

 space and extending beyond the earth's orbit, but none of 

 them, so far as I know, has suggested the interior of the sun 

 as the place from which the Zodiacal substance has been 

 ejected. 



That cometary and meteoric matter may have contributed 

 to the volume of discrete bodies surrounding the sun and 

 extending to some distance within the orbit of Mercury has 

 some degree of probability in its favour, but the extreme 

 tenuity of the outermost parts of the Zodiacal substance, 

 together with its immense distance from the central body, 

 appears to me to be better accounted for on the supposition 

 of its consisting of the lighter elementary substances in a 

 state of extreme sub-division ejected during solar eruptions, 

 as in the instance of the ejection of enormous masses of 

 hydrogen observed by Young which 1 have already 

 adduced. 



LXXIV. Limitations imposed by Slip and Lnertia Terms upon 

 Stokes's Law for the Motion of Spheres through Liquids. 

 By H. D. Arnold, Fellow in Physics in the University of 

 Chicago*. 



RECENT investigations have shown that Stokes's law 

 for the motion of spheres in fluids is less general in its 

 applicability than has hitherto been assumed. The importance 

 of these investigations has lent interest to a determination of 

 the limits within which the law may be applied, and of the 

 error introduced when these limits are exceeded. Liquids 

 of moderate viscosity lend themselves most readily to deter- 

 minations of this kind, since in them spheres of considerable 

 size will move with easily measurable velocities, and 

 especially since the viscosities of these liquids may be very 

 simply found with a high degree of accuracy. It is the 



* Communicated bv Prof. R. A. Millikau. 



