280 KELVIN 



If you do not see it, it is because your eyes are not sensitive 

 enough, but a little training will give them the needed sen- 

 sitiveness. If you cannot see it in this way try another 

 method. Look into a pail of water with a black bottom; or 

 take a clear glass dish of water, rest it on a black cloth, and 

 look down at the surface of the water on a day with a white 

 cloudy sky (if there is such a thing ever to be seen in Phila- 

 delphia). You will see the white sky reflected in the basin 

 of water at an angle of about fifty degrees. Look at it 

 with the head tipped on one side and then again with the 

 head tipped to the other side, keeping your eyes on the water, 

 and you will see Haidinger's brushes. Do not do it fast 

 or you will make yourself giddy. The explanation of this 

 is the refreshing of the sensibility of the retina. The 

 Haidinger's brush is always there, but you do not see it be- 

 cause your eye is not sensitive enough. After once seeing 

 it you always see it ; it does not thrust itself inconveniently 

 before you when you do not want to see it. You can also 

 readily see it in a piece of glass with a dark cloth below it, 

 or in a basin of water. 



I am going to conclude by telling you how we know the 

 wave-lengths of light, and how we know the frequency of 

 the vibrations, and we shall actually make a measurement 

 of the wave-length of yellow light. I am now going to show 

 you the diffraction spectrum. 



You see on the screen," on each side of a central white bar 

 of light, a set of bars of light, of variegated colours, the first 

 one on each side showing blue or indigo colour, about four 

 inches from the central white bar, and red about four inches 

 farther, with vivid green between the blue and the red. That 

 effect is produced by a grating with 400 lines to the centi- 

 metre, engraved on glass, which I now hold in my hand. The 

 next grating that we shall try has 3,000 lines on a Paris inch. 

 You see the central space and on each side a large number 

 of spectrums, blue at one end and red at the other. The 

 fact that, in the first spectrum, red is about twice as far 

 from the centre as the blue, proves that a wave-length of 

 red light is double that of blue light. 



* Showing the chromatic bands thrown upon the screen from a diffrac- 

 tion grating. 



