December i6th, ipo2.] PROCEEDINGS. xi 



of polishing. This explanation would account for the irregular 

 reflection of the rim, consisting of bright spots and patches, 

 by the slight crumpling which would be caused by the stress of 

 solidification ; also for the concavities representing the compara- 

 tive softness of the inner parts of the thick area. 



Dr. Charles H. Lees showed a large magic mirror, from 

 which an excellent reflection was obtained, and an interesting 

 discussion ensued on the subject of these mirrors. 



Mr. R. W. Ellison exhibited a series of eggs of the Common 

 Guillemot {U'rm troile)^ showing great variety in coloration and 

 design of markings, eggs of various shades of green, blue, yellow, 

 brown and red being prominent. The light coloured eggs pre- 

 dominate, fully 90 per cent, being shades of green and blue 

 variously marked with brown or black, the dark shades of brown 

 and red being only occasionally met with. The eggs are not pro- 

 vided with a protective colouring, but their shape prevents them 

 from rolling off the narrow sloping ledges on which they are 

 deposited. The birds breed in colonies on rocky headlands, and 

 each bird lays one egg on the ledges of the cliffs. If the 

 egg be taken or destroyed whilst it is fresh, a second egg is 

 laid a fortnight later, but, should incubation have commenced, 

 no more eggs are laid during the season. While incubating, 

 the bird sits facing the cliff and holding the egg between 

 its legs, but, should the egg be removed, the bird will 

 then turn and face the sea. The eggs of each individual bird do 

 not vary in the ground colour, that is to say, a bird that has once 

 laid a green egg will not lay eggs of any other colour. It has also 

 been noticed that each bird invariably returns to the same spot 

 on the cliff" to lay its eggs. 



The following paper was read : — 



"The Graphic Computation of Lenses." 



By C. E Stromever, M.Inst.C.E. 



\Abstract^ 

 Having drawn attention to a remark in Lummer's " Optics " 

 to the effect that the computation of oblique rays which do not 



