130 Mr. S. Bidwell. On the Recurrent [June 7 y 



piece with a large field. I believe that, after sufficient practice, this 

 would he found the best possible method of observation ; but it is 

 exceedingly difficult to keep the eye absolutely steady, and untrained 

 observers never succeeded in seeing the looked-for phenomena at 

 all.* Since it seemed desirable that my own observations should be 

 confirmed by others, I abandoned the telescope and the ground glass, 

 and by means of the lens N" focussed the reflected image of the 

 aperture upon a white screen, R. The diameter of the projected 

 disk of coloured light was about 1'5 cm., and that of the approxi- 

 mately circular path which it described, 30 cm. To aid in steadying 

 the eye, a spot of luminous paint, upon which the gaze might be 

 directed, was applied at the centre of the circle. With this arrange- 

 ment, almost any one can see the ghosts without the smallest 

 difficulty. 



When the mirror turns once in 1^ sees., the ghost or recurrent 

 image appears about 50 behind the coloured disk, the corresponding 

 time interval being one-fifth of a second. Exact measurement is, 

 however, not easy, and it is probable that the interval is not quite 

 the same with light derived from different portions of the spectrum. 

 The ghost appears to be circular in form, its diameter being generally 

 rather less than that of the original. The colours of the recurrent 

 images, as specified below, have all been observed by several persons r 

 and, except as to those at the extreme limits of visibility, all the 

 observations were in agreement. 



Experiment 1. 



Spectrum colours. Recurrent colours. 



Extreme violet No perceptible image. 



Middle violet A pale image, variously described as grey, yellow, and 



greenish-yellow. 



Dark blue Feeble violet. 



Light blue Brighter violet. 



Middle green Bright violet. The image is more conspicuous with 



green light than with any other. 



Greenish-yellow Blue. 



Orange-yellow Bluish-green, 



Orange Dark bluish-green. 



Orange-red., . , Very dark bluish-green. 



Red No image at all, however bright the red was made. 



The violets all appear to my own vision slightly redder than the 

 violet of the spectrum. 



The following experiment was then made. 



* If a telescope is used, the mirror must be silvered on its outer surface, and in 

 the air of a laboratory is quickly tarnished. 



