of cubical expansion. 



327 



The cathetometer readily detects and measures a motion of 

 •^jj-inm. ; so it appears that a change of temperature of T ^° 

 causes a change in level of water in the tube which can observed. 

 The thermometer, however, was read only to ■^ w °. 



It remains to examine into the effect of the glass tube at- 

 tached to the metallic vessel. This tube is necessary in order 

 to observe the level of the water, but to get a rigorously correct 

 observation the vessel and tube should be of the same substance. 

 The effect of replacing the glass tube by a similar one of the 

 same material as that of the vessel is readily calculated and is 

 found to be a quantity that may be neglected. Thus in a brass 

 tube 50 mm. long and 203 mm. in diameter the level of the 

 water at the higher temperature of 16° will stand 014 mm. 

 lower than in a glass tube of the same dimensions, and at 21 0, 5 

 the water will be - 05 mm. lower in a zinc tube than 

 in one of glass. 



The coefficient of cubical expansion of glass was 

 determined by means of a spherical vessel of 12 cm. 

 diameter with a tube of 2*5 mm. interior diameter. 

 The level of the water in this vessel was the same at 

 ll°-75 and at 0°. The volume of water at 11° -75 is 

 1-00U3 and T T °.f|- = '0000255 the coefficient of cubical 

 expansion of this kind of glass. Kopp gives '000024 

 and "000026 as the coefficients for a similar soda glass. 



The coefficient of expansion hydrometer was used 

 to' determine the coefficients of expansion of brass 

 and of hard rubber, or ebonite. The hydrometer, fig. 

 3, was made of the same kind of brass as that form- 

 ing the cylinder used in the determination of this 

 coefficient. The cylinder of the hydrometer is 25 - 5 

 cm. long terminated by cones. Its diameter is 7*3 

 cm. The stem of the hydrometer is a brass tube of 

 2 mm. exterior diameter. The hydrometer of hard 

 rubber is a cylinder 21*5 cm. long and 6*5 cm. in di- 

 ameter with a stem 2 mm. in diameter. To float 

 these hydrometers in water at 0° and at a higher 

 temperature, a pointed rod, as shown in the figure, 

 was used. This point, by just touching the surface 

 of the water, showed the depth of flotage. These 

 hydrometers were made to float in water at 0°, with \ / 

 their index-points just touching the surface of the 

 water, by loading them with fine shot. The water was then 

 slowly heated and it was found that the brass hydrometers 

 floated to the same depth at 0° and at 15°*85, and the hard rub- 

 ber hydrometer had the same depth of flotage at 0° and at 

 38°-55. 



Am. Jour. Sci. — Third Series, Tol. XL, No. 238.— Oct., 1890. 

 21 



