1 83 1 .] On Sounds from heated Metals. 3 1 1 



be again developed, and it is easy in this way to separate many 

 colours from each other. The experiment in illustration of 

 Newton's theory of colour, by painting the head of a top and 

 spinning it, is well known ; by the means just described the 

 experiment can be still further extended, and the colours sepa- 

 rated one from another, even while the whole system remains 

 in motion. 



The combination of other forms than wheels by the apparatus 

 described, page 294, produces very beautiful effects. The 

 application of colours here also is so evident as to need no 

 illustration. The variation of the proportion of the interval to 

 the remaining pasteboard causes many curious appearances, 

 especially when the shadows produced in sunlight are observed. 



Since the printing of the paper, a friend has referred me to 

 the article ' Animalcula ' in Brewster's Encyclopaedia, where 

 an opinion on the appearance of these creatures is given, nearly 

 the same as that I have ventured. Speaking of the opinions of 

 those who suppose them to be true revolutions^ it is said, " Yet 

 notwithstanding our respect for the skill and talents of such 

 renowned naturalists, we cannot deny that we think the pro- 

 duction of the vortex is more probably effected by the simple 

 motion of the fibrilla that it may ensue from their rapidly 

 bending in regular or alternate succession, or by some analogous 

 means." 



Trevelyan's Experiments on the Production of Sound during 

 the Conduction of Heat*. 



[Read Friday evening, April 29, 1831.] 



MR. TREVELYAN had remarked that when a heated poker was 

 laid down upon a table, so that the knob rested upon it, whilst 

 the hot part was supported by an interposed block of cold lead, 

 regular musical notes were frequently produced. By extend- 

 ing the experiments, he found that a better form than that of 

 a poker might be used for the hot metal: a piece of brass 

 about four inches long, one inch and a quarter broad, and half 

 an inch thick, should have a groove of one-eighth of an inch in 

 width, formed down the middle of one of the broad faces, and 

 then that face bevelled from the edges of the groove on each 



* Quarterly Journal of Science, 1831, ii. 119. 



