AND RATIO OF THKIR ATOMIC WETOTITS. 9 



COMPUTATION OK THE OBSERVATION AT I 2.30 P.M. 



Weights ABCDEFGHJ H2SM gr. 



Reduction to vacuum — i'7 g""- 



Reduction of weights to temperature, 17.24" C — .1 1 gr. 



Corrected weight of brass used as a sinker 8424.36 gr. 



Cage and pan 178-32 gr- 



Globe No. I, corrected to va( uuni 1015.22 gr. 



Sum of weights, left pan of balance 9617.90 gr. 



Weights on right pan, reduced to vacuum 35 1.09 gr. 



Weight of water displaced 9266.81 gr. 



Multiplving by T.001237, we have volume at 17.24" 9278.27 cc. 



Reduction to o" for expansion of glass — 4-47 cc 



Volume at o°C 9273.80 cc. 



In the same way, we get for the other weights and temperatures given above, 



12.50 P.M. 9273.79 cc. 



2.15 P.M. 9273.80 cc. 



3.15 P.M. 9273-79 CC. 



4.30 P.M. 9273-79 CC. 



6.00 P.M. 9273.79 CC. 



These deterniiuatious are recorded as one I'esnlt: all the results for this globe are 

 given in the table of data concerning the capacity of globes. 



IIYDKOSTATIC WEKUUNO OF THE l.,M:<iKK (JLOBKS. 



In the case of the largei' globes, tlie manipulation in the hydrostatic weighings 

 was slightly different; the cylinder filled with water in which the globe was 

 immersed was too heavy to be conveniently moved into and out of the case on 

 wliich the balance had been placed. The cylinder for the larger globes was accord- 

 ingly placed permanently in a closet about three metres square which was built 

 inside a basement room of tolerably constant tempei'ature. It was surrounded with 

 a non-conducting case. ( )ver it was placed the balance with a clear space of three 

 quarters of a metre beneath it. Further, the volume of water needed was now so 

 much that the removal of it every time a globe was immersed became tedious. So 

 a triangular platform was attached to three brass I'ods, and these to cords passing 

 over pulleys in the ceiling, so that the platform with any load could be raised or 

 lowered at will. The cords wei-e arranged to pass before, behind, and to the left 

 of, the balance without interfering with it. This platform being raised, the pan 

 and its weights were arranged and the globe and its collar were attached to it. 

 The platform was then lowered, and the globe sank to the bottom. The latter was 

 then raised a little and suspended l)y a wii-e from the balance, and weighed as 



l)efore. 



In the other method, the water was stirred by poui'ing, which \vas tedious. 



