Polarization of the Light of the Sky. 515 



object of physical theories of this kind is to account for the ob- 

 served facts by an explanation which employs only the ordinary 

 principles of mechanics ; so that any communication of energy 

 between the aether and material particles must be effected by 

 the action and reaction of pressures arising out of relative 

 motion; otherwise the explanation is of the kind ignotum per 

 ignotius. Now if we make in so many words the hypothesis that 

 the density of the sether is the same within as without all bodies 

 without further restriction, the effect is to suppose that the ulti- 

 mate particles of matter and the sether mutually interpenetrate 

 each other, i. e. that they can occupy the same space without in- 

 terfering with each other. This is, as far as I can see, not a phy- 

 sically conceivable hypothesis. And when we turn to the mutual 

 action, we have to make some additional hypothesis and intro- 

 duce forces originating in a different manner from the pressures 

 which act between any two bodies not mutually interpenetrable. 

 Any hypothesis, for instance, which supposes mutual interpenetra- 

 tion excludes such analogies as that of Lord Rayleigh above re- 

 ferred to, in which the ellipsoid and surrounding fluid are cer- 

 tainly not supposed interpenetrable. It seems to be that the 

 hypothesis needed is something of this sort — that material bodies 

 are built up like a complex network in space, the interstices of 

 which are considerable as compared with the space occupied by 

 the material rods or points of which they consist. The relation 

 to the aether may be illustrated by supposing such a network 

 built up of rods like very fine needles ; the whole could then be 

 pushed with ease into a mass of jelly. The jelly would separate 

 before and close again after the rods as they passed through it. 

 The separation of the jelly by the rods is not inconsistent with 

 its being substantially incompressible ; but the result would be 

 that the sether would be parted on each side of a solid body 

 moving through it to an extent depending on the ratio of the 

 space occupied by the matter to that occupied by the sether within 

 the body. It seems to me that some such hypothesis as this is 

 necessary for any explanation of the facts of radiation and absorp- 

 tion ; and so far as I can see, it would not be incompatible with 

 the application of Green's theory or of that proposed by Lord 

 Rayleigh. It is fully consistent with Lord Rayleigh' s treatment 

 of the problem of the first stage of TyndalPs clouds. It also 

 seems to me to afford an explanation of the phenomenon of re- 

 sidual blue. 



It is only necessary to admit that the size of the particles in 

 the first stage is small, not only with respect to the wave-lengths, 

 but also with respect to the amplitudes of the incident light. 

 Consider then the stage at which the mean diameters of the 

 particles are equal to the amplitudes of the incident light. (In 



2L2 



