4«i$ Astronomical and Nautical Collections, 



primitive undulation, being perpendicular to its surface, the 

 motions of the particles of ether in this direction must be 

 more considerable than in any other ; and the rays depend- 

 ing on these motions, if separately considered, would be so 

 much . the. weaker as they deviated the more from this di- 

 rection. 



The investigation of the law by which their intensity would 

 be governed, according to their direction, as derived from 

 any separate centre of agitation, would certainly be of very 

 difficult investigation : but happily we are not obliged to 

 determine this law, for it is easy to see that when the incli- 

 nation to the perpendicular is considerable, the effects of the 

 different rays must very nearly destroy each other : so that 

 these rays, which sensibly affect the quantity of light re- 

 ceived at each point P, may safely be regarded as being equal 

 in intensity. - > ^c-t 



When the centre of agitation has undergone a condensa- 

 tion, the expansive force tends to urge the molecules in every 

 direction ; and if they do not perform a retrograde motion, 

 it is only because their initial velocities forwards destroy 

 those which the expansion of the condensed fluid would 

 otherwise generate backwards : but it does not follow from 

 this that the agitation can only be propagated in the direc- 

 tion of the initial velocities ; for the expansive force in a 

 perpendicular direction, for example, will combine with the 

 primitive impulse without any diminution of its effects. It 

 is obvious that the intensity of the undulation thus produced 

 may vary much at the different points of its circumference, 

 not only from the nature of the initial impulse, but also be- 

 cause the condensations are not subject to the same law on 

 every side of the centre of the agitated part [?] . But the 

 variations of the intensity of the derivative undulation must 

 necessarily be subjected to a law of continuity, and may con- 

 sequently be considered as insensible in a very small angular 

 interval, especially in the neighbourhood of the perpendicular 

 to the surface of the primitive undulation ; for the initial 

 velocities of thq. .iiaplecules, referred to any given direction, 

 being proportional to the cosines of the angles made by that 

 direction with the perpendicular, these results vary rauoh 



