24« On certain Conditions under which 



effect due to diiFusion and absorption. Yet as, in order to 

 require time for its propagation, it must pass progressively 

 from point to point throughout the whole course of its journej', 

 it seems clear that it must consist of, or involve, matter in 

 some form, however subtile and rethereal ; and the analogy 

 of all fluids would warrant the conclusion that there must be 

 an actual loss of force when the moving power becomes 

 more remote. If this be so, a ray of sun-light must consist 

 of a series of rays between certain limits differing in their 

 powers, and possibly in their properties; for when we consider 

 the intimate dependence of the refraction of light, for in- 

 stance, upon the density of the media through which it has to 

 pass, and of its several forms of polarity upon the physical 

 composition of the substances which produce it, who can say 

 that these phgenomena are beyond the reach of such a source 

 of influence? We know so little of the mysteries of polari- 

 zation, so little beyond its outward and visible effects, that it 

 is impossible to conclude that it has no relation to such a cir- 

 cumstance ; and although in both the above branches of op- 

 tical inquiry I am well aware that much might be answered 

 in respect of observations upon light from sources where no 

 such cause of influence could exist, still so far as they have 

 been made without a view to this particular point, they do not 

 amount to a demonstration that in no case could any differ- 

 ence arising from it be detected. 



There is another point which deserves consideration. Al- 

 though the diameter of the sun subtends a sensible angle at 

 the eye, still the distance is so great, and consequently that 

 angle is so small, that we may assume the rays, even from its 

 extremities, to be parallel, for the purpose of general reason- 

 ing. In that case it is evident that the series of increased 

 distances to be traversed by them will be represented by the 

 versed sines of the angles whereby they are removed from the 

 centre of the visible hemispherical surface. Now if we sup- 

 pose undulations or vibrations of any kind as the cause of the 

 sensation of light, a supposition to which we are led by many 

 optical facts, it is obvious, that while, in this series, there is a 

 certain number of points which correspond exactly to those 

 of the departure of successive waves, there must be also an 

 intermediate number which would tend to neutralize the first, 

 and to produce the well-known effect termed intafcrencc. 

 Indeed, the series of points of propagation being unbroken 

 over the whole curved surface, and mathematically continuous, 

 it is diflicult to conceive, upon the hypothesis of successive 

 waves, how this effect can fail to take place, so immediately 

 and absolutely, without an interval between one and another, 



