DENSITIES OF OXYGEN AND HYDROGEN 



B. EFFECTIVE WEIGHT IN WATER AT l8 C. KEUL'CEU TO WEICHT IN VACUUM : GRAMMES. 



The probable enoi' of a single determination in Table A is 10, and in Table 

 H is 4.1), milligrammes; that of a mean value in A i^ (i, and in B is 2.;'), niilli- 

 graranies ; that of the sum of all the large weights in A is ■_'(•, and in l*> is l(i, mil- 

 ligrammes. It is obvious that the values of the weights were sutticientiv \vf^ll 



determined. 



-nYDROSTATIC WETfUIlA'fi OF GLOBES. 



The manipulation of the hydrostatic Aveighing of globes varied slightly with 

 the size of the globes. 



For the smaller globes nundiercd I and L>, the ari'anwmcnt sIkiwh in Fie 2 

 was used. The globe is in a cojtper tyliiidei' having tubulatures for si.\ thermome- 

 ters. Around the neck of the globe is placed a brass collar, shown at <■ in Fig. .'?. 

 To this is suspended a pan a by means of copjier ribbons b h. Distilled water is 

 [loui'ed into the cylinder to a depth of perhaps twenty centimetres. The jian is 

 laid in the bottom of the cylinder and loaded with the required weights, care being 

 taken to remove air Imlibles. The globe and its collar are adjusted in position, 

 and the cylinder is filled with water without producing air bubbles. The cylinder 

 is then set in the noii-cnnducting case on which the balance stands, Vvj:. i», and 

 the globe is siispeiuli'd from nm- of its pans by asmall wire. The thermometers are 

 thi'ust far enough through the tuiiulatures, and the case is shut. After a time, the 

 weights needed to produce equilibrium aie determined, when the case is ojiened 

 and the thermometers are read. 



