6 Ohservatiotn on Hefraclion. 



gives a detail of the experiment of placing a piece of money at 

 the bottom of a tumbler partly filled with water ; and from his 

 observing that the piece was seen when the tumbler was held 

 below, on a level with, and above the eye, he concludes that 

 in each case it is seen by a reflected image formed perpendi- 

 cularly over it on the sm-face of the water. 



Before entering into any examination of the weight of Dr. R.'s 

 arguments, I shall here mention two experiments which I think 

 are of themselves quite sufficient to set at rest the whole of his 

 rea-soning on the subject. 



Exp. 1. — To jireclude the possibility of the surface of the 

 water in the tunibler becoming a reflecting surface, I covered 

 it over with a circular piece of dry flannel, with a small semi- 

 circular hole cut out of its edge. On holding the tumbler be- 

 low the level of the eye, the half-crown was seen through the 

 opening in the edge, in the same manner as before it was co- 

 vered ; but on holding it above the eye no image whatever 

 coukl be seen. I would here ask Dr. K., how it happens that 

 the reflected image is destroyed in the one case and not in the 

 other. 



Exp. 2. — Having formed a small tube ^.^^ 



of pasteboard with an angular bend in it [ 'd^^::^^^^^\ 



at D, as in the figure, so that nothing - ^^y^- b 



could be seen through it in the open air, /' 



I placed the end C on the bottom of the c 



tumbler A B, the part C D making an 



oblique angle with the surface of tlie water, and having tlie 

 point D exactly in that surface. By holding a candle under 

 the point C, and looking through the tube from the other 

 end E, the bottom of the tumbler was seen quite distinctly, 

 the tube at the same time appearing nearly straight. When 

 a straight tube was used and held in the same direction C D, 

 nothing whatever could be seen through it. From this ex- 

 periment I may also draw Dr. Reade's weighty inference, that 

 " ifo see is to believe " but that to see an^ object through a 

 bended tube in the above manner, is to believe that the rays of 

 light in their progress from the object to the eye follow the 

 direction and bend of that tube ; or, in other words, they are 

 refi-acted in passing from water into air. 



^ What appears to me to have led Dr. R. astray in most of 

 his reasonings is the singular opinion he seems to entertain, 

 (although he has not expressly mentioned it,) that the rays of 

 light do not proceed in all directions from every point of an 

 object, but that they all go on in one particular direction pa- 

 rallel to each other. For example, in his first experiment, 

 he says, " Let us examine this experiment according to the 



received 



