1892.] Light seen as if emanating from Flames. 71 



hope has not been fulfilled, and I had offered to the Secretaries a 

 communication describing my recollection of what my brother had 

 told me, when his son found a memorandum of date 18th October, 

 1891, and a little book of notes of date 29th December, which tell 

 the story better than I could have told it, and which, therefore, 

 though not completed in proper form for publication, I now give in 

 the unfinished form in which they have been found, with only a some- 

 what more clear drawing, and description of drawing, substituted for 

 the rough sketch found in his note of date October 18, 1891. 



Proposed probable Paper for the (?) Society, by J. T., " On the Nature 

 and Origin of certain Appearances of Beams of Light as if emanating 

 from Candle or Lamp Flames." 



Description of the Drawing. 



[The drawing represents a vertical section of the eye, eyelids, and 

 watery prismoids,* through FF', the axis of the eye. The large num- 

 ber of parallel lines outside represent rays of light coming from a 

 flame several feet or yards away in the direction of F', to the eyelids, 

 the prismoids, and the undisturbed outer surface of the cornea between 

 the prismoids. The lines within the eye below FF' represent the con- 

 vergence to F, the image of the flame, of those of the external rays 

 from the flame which fall on the undisturbed portion of the surface 

 of the cornea. The lines within the eye above FF' represent rays dis- 

 turbed by the prismoid of the upper eyelid which, incident on the 

 retina at bbb, give the perception as if of light coming from without 

 in the direction of the dotted lines outside the eye. It is this per- 

 ception that constitutes the appearance of the downward beams or 

 ladders of light, due to the prismoid of the upper eyelid. The rays 

 disturbed by the prismoid of the lower eyelid, in the position repre- 

 sented in the diagram, are all stopped by the lower part of the iris. 



Looking now at the diagram, we understand perfectly that if, with 

 the eyeball and flame unchanged, the upper eyelid be gradually raised 

 a little, the uppermost of the rays coming inwards from the prisntoid 

 will fall on the upper part of the iris and will be stopped by this 

 screen. Thus, the length of bbb upwards from F is diminished, until 

 all the beams from the prismoid are stopped by the iris, and the 

 length of the apparent beams below the flame correspondingly 

 diminishes to zero. When the upper eyelid is wide open the flame 

 is seen without any appearance of the beams below it. We also 

 understand readily from the diagram how, if the lower eyelid is lifted 

 a little without any change in the position of the upper eyelid, beams 

 both above and below the flame are seen. We also conclude that if, 



* The refracting watery liquid in tlie entrant corner between lip of eyelid and 

 cornea may be called the prismoid or liquid prismoid. 



