of the prismatic Colours to beat and illuminate Objects. 279 



well, and the smoke was not disordered ; but, when all the tube 

 was open, the pale green glass cracked in a few minutes. 



No. 9. Placing now a dark green before a smoked green, I 

 saw the sun remarkably well. In this experiment, I had made 

 a difference in the arrangement of the apparatus. The cracking 

 of the glasses, I supposed, might be owing to their receiving 

 heat in the middle, while the outside remained cold ; which 

 would occasion a partial dilatation. I therefore cut them into 

 pieces about a quarter of an inch square, and set three of them 

 in a slider, so that I could move them behind the smoked glass, 

 without disturbing it. After looking about three or four minutes 

 through one of them, I moved the slider to the second, and then 

 to the third. This kept the glasses sufficiently cool; but the 

 disturbance of the alterations proved hurtful to vision, which 

 requires repose ; and, if perchance I stopped a little longer than 

 the proper time, the glass cracked, with a very disagreeable 

 explosion, that endangered the eye. 



No. 10. Two dark green glasses, both smoked, that a thinner 

 coat might be on each ; but the smoke still contracted blisters, 

 though less dense than before. 



No. 11. To get rid of smoke intirely, I used two dark green 

 glasses, two very dark green, two pale blue, and one pale green 

 glass, together. Distinctness was wanting ; nor was light suffi- 

 ciently intercepted. 



No. 12. A dark green and a pale blue glass, smoked. The 

 green glass cracked. 



No. 13. A pale blue and a dark green glass, smoked. The 

 blue glass cracked. The eye felt no sensation of heat. 



No. 14. Two pale blue glasses, one smoked. The first glass 

 cracked. 



mdccc. O o 



