210 Don Bu stamen te on a new Graxnmeter. 



as there is no necessity for knowing the weight of the mercury 

 contained in the tube of the barometer for ascertaining and 

 comparing the various pressures of the air. It is sufficient, then, 

 for us to fix with precision the point of the scale at which 

 the water stands, before loading the instrument, and that to 

 which it sinks in consequence of the loading applied : To this 

 alone is its use reduced. 



9. Suppose that, being loaded with an additional weight, the 

 level of the water reaches exactly the zero of the scale, if in 

 this state we put on the upper plate a fragment of calcareous 

 spar for example, the weight of which produces an immersion, 

 so that the level of the water should mark the division 54, 

 this number shows us the weight of the fragment weighed in 

 air (§ 7.) If we change then the fragment to the lower plate 

 the immersion only reaches to 84, and this number shows us 

 the weight of the same fragment weighed in water. Then the 

 difference of 20 between these two numbers is exactly the 

 weight of the volume of water dislodged by this substance, or 

 the weight lost in the second operation. There remains only to 

 divide 54, the weight of the fragment in air, by 20, which is that 

 which is lost in water. The quotient, 2.7, shows us the specific 

 weight of the calcareous spar; and this simple method must be 

 followed for all other bodies. 



10. It is easily perceived that the additional weight may be 

 greater than we supposed, without its altering the data, be- 

 cause if, instead of sinking the instrument to zero, it should be 

 sunk to the division 8 for example, then the same fragment of 

 calcareous spar would have immerged it not only to 54 but 

 to 62, and always its weight in air would be the same as before, 

 that is to say, 62 — 8 = 54. The same happens with the weight 

 in water. In the second instance the immersion would not be 

 to 34, but would rise to 42, and the loss would be equal to the 

 former, that is 62 — 42 = 20 ; and this is one of the advanta- 

 ges of the instrument, that it is not requisite to bring it to any 

 fixed point, but only to observe the divisions of the scale which 

 mark the level of the water, as already said (§ 8.) 



11. If we take away from the upper plate not only the sub- 

 stance that has been weighed, but also the additional weight, 

 the instrument rises till it leaves the lower plate above the 



