On Maxwell's Stress Theory. 491 



number of thin lenses in contact is the sum of their individual 



convergences, we immediately have the interesting result 

 that for combinations of thin lenses in contact, without 

 diaphragm separated from the combination, the curvatures of 

 thy field< may be added together just as with convergences. 

 We thus have 



Li 



where F = wF. 



For such a combination the mean curvature therefore is 



F 

 2F-I-- — . or 2p if the Petzval condition is satisfied, aud 



under these circumstances B l9 ~R 2 , and R 3 = F, 2F, and 3F 

 respectively, while the curvature corresponding to the 

 " astigmatic difference " is always 2p independently of the 

 materials of the lenses. 



This example will serve to show how easily and directly 

 aberration problems may be solved by physical methods, and 

 the writer proposes to show in another paper that the whole 

 theory of aberrations in refracting systems may be similarly 

 treated with advantage. He ventures to hope, however, that 

 enough has been said in this paper to convince everyone 

 that not only is there no necessity for the abandonment of 

 curvature methods at any stage in optical work, but that there 

 is every advantage in retaining them throughout. It will 

 also be obvious that they lend themselves to combining the 

 study of diffraction and image formation, which should lead 

 to valuable new result-. 



XLIV. On Maxwell's Stress Tlieory. ByV. Bjerknes *. 



MAXWELL considered his theory of the stress in the 

 dielectric medium as very important. But, on the 

 other hand, he did not regard it as complete. His own 

 words plainly prove both assertions t : — 



" It must be carefully borne in mind that we have made 

 only one step in the theory of the action of the medium. 



* Communicated by the Author. 



t Maxwell, 'Electricity and Magnetism,' 2nd edition, vol. i. p. 154. 



2 K 2 



