296 ON LIGHT. 



not very large, they will require a magnifier to be well 

 seen, their diameters being in that case very small ; .but 

 with a lens of 20 or 30 feet focal length it is consider- 

 able, and the rings may be seen, and their diameters 

 measured, with ease. Now it is found that these diame- 

 ters, for the. first, second, third, &c., dark rings in order 

 (reckoned from the centre), are not in the proportion of 

 the numbers i, 2, 3, &c, but of the numbers i, 1*414, 

 1732, 2-000, &c., which are their exact square roots, 

 giving to their system the appearance represented in the 

 preceding diagram ; and this is exactly the progression of 

 distances from the point of contact measured on the 

 surface of the plane glass which correspond to the 

 series of perpendicular distances between it and the convex 

 spherical surf ace of the upper glass in the proportion of the 

 arithmetical series, as may be seen in Fig. 7. 



(79.) So far, then, the Newtonian hypothesis affords a 

 satisfactory account of the facts ; in all, that is, but that 

 one particular already adverted to. This, however, must 

 be considered as conclusive against it ; while, on a con- 

 sideration of the whole case, there remains outstanding 

 this strange fact that at certain distances between two 

 partially reflecting surfaces, forming a regular arithmetical 

 progression from nil upwards, the portion of a beam of 

 light reflected from the second, after passing back 

 through the first, so far from augmenting the first reflected 

 light, annihilates it, and furnishes us with an instance 

 (which is, as we shall see hereafter, not the only one) of 

 the combination of lights creating darkness I 



(80.) The question now arises, Will the undulatory 



