264 THE GRAMMAR OF SCIENCE 



liberty to first consider it as of one nature and then as of 

 another. But in doing so it is evident that we are leaving 

 room for a wider conception which will resume both 

 phases of phenomena at once, and will not lead us into 

 logical contradictions if both phases have to be dealt with 

 in the same investigation. Thus, if the ether as a perfect 

 fluid enable us to describe atoms by its types of motion, 

 and the ether as a perfect jelly enable us to describe the 

 radiation of light, it is clear that when we treat the atom 

 as a source of light-radiations, we may get into serious 

 confusion by the conception that the ether is at the same 

 time a perfect fluid a7id a perfect jelly. We are compelled, 

 indeed, to try and find some reconciliation between these 

 two conceptions. If we turn to perceptual experience for 

 a suggestion, we may note that water is the principal 

 component of jelly, and may, by the addition of more or 

 less gelatinous material, be stiffened to a jelly of any 

 consistency. In the like manner we can conceive a series 

 of perfect jellies formed, ranging in their resistance to 

 slide, from the perfect fluid, through all stages of viscosity, 

 up to the perfectly rigid body. We might, then, out 

 of this series of jellies choose one which, for sliding strains 

 of a certain magnitude, was sensibly a perfect fluid, while 

 for smaller strains, such as are involved in the theory of 

 light-radiation, it would act as a perfect jelly. This is 

 the solution propounded in 1845 by Sir George G. 

 Stokes,^ and it may be termed the jelly-theory of the 

 ether. The jelly- theory of the ether has undoubtedly 

 been of value in simplifying many of our conceptions of 

 physical phenomena, but how far it can be reconciled with 

 any system of ether-motion as a basis for the prime-atom 

 yet awaits investigation.^ 



1 Mathematical aiid Physical Papers, vol. i. pp. 125-29, and vol. ii. pp. 

 12-13. The present writer considers, however, that there is a difference in 

 quality as well as in degree between a viscous fluid and an elastic medium. 

 The complete difference in type between the equations of a plastic solid and 

 a viscous fluid is sufficient evidence of this. In the former case, any shear 

 above a certain magnitude produces set ; in the latter, any shear whatever, if 

 continued long enough. 



2 For example, Lord Kelvin's vortex atom would hardly be a possi- 

 bility. 



