the Thermal Expansion of' Marble. 63 



a wooden beam compass, kept as nearly as possible at one tem- 

 perature, was likewise compared with the marbles. 



In each slab, at the distance of 31.5 inches, were inserted two 

 brass pins, one of which had a minute hole drilled in its centre, 

 for the purpose of receiving one point of the compass, while the 

 face of the other was smoothed, to allow of faint traces being 

 made on it with the remaining point. 



The slabs and the glass compass were placed in a tin trough, 

 and surrounded with broken ice, a sufficient quantity of water 

 being poured in to perfect the communication between the ice 

 and the marble. At the end of an hour, when the marble miffht 

 be supposed to have attained the temperature of melting ice, a 

 sufficient space was cleared of icy fragments to allow of motion 

 to the glass compass, without raising it above the surface of the 

 surrounding fluid, and faint traces were made with both com- 

 passes on the smooth pins in each of the slabs. 



The ice and water were then removed from the trough, and 

 their place was supplied with hot water, which was kept boiling 

 at 211° Fahr. for ^n hour. At the end of that time, traces were 

 again made with each of the compasses, the glass compass lia- 

 ving remained completely immersed in the boiling fluid, and the 

 wooden one having been hastily brought from an adjoining apart- 

 ment, whose temperature had not varied. However, on account 

 of the short but unavoidable delay occasioned by the removal of 

 the upper slab, the wooden beam must have suffered some slio-ht 

 increase of temperature and of length before the trace was made 

 on the pin in the Carrara marble. 



The distances between the traces were then examined with a 

 microscope, and mea-sured by means of a silver feather-edge, 

 wliose markings could be depended on to the 3000th part of an 

 inch. 



The total expansions given by the wooden compass were 

 jo'c^h of an inch for the Lucullite, and ^H^th for the Car- 

 rara marble, on a length of 31-5 inches, and for a change of 

 temperature from 32° to 211° Fahr.; that is, for the entire 

 change of 180°, an expansion in the black marble of -000350, 

 and in the white of '000837. 



On examining the traces made by the glass compass, it was 

 found that the Liicullitc marble had expanded less than had 



