160 



PROCEEDINGS OF THE AMERICAN ACADEMY 



greatest efficiency when placed near the top of the space to be pro- 

 tected. The tin box was itself enclosed in a cupboard, but not other- 

 wise protected ; and the balance case was surrounded by curtains, in 

 order to shield the beam from radiation. 



With the apparatus so arranged, it was found possible to obtain 

 most satisfactory and concordant results to tenths of a milligram 

 when the change in the temperature of the balance-room was not 

 very rapid ; but any sudden changes produced by artificial heating 

 would cause slight currents of air in the interior of the case, whose 

 effect became very sensible, but whose influence we were able to 

 eliminate. The best series of results, however, (the second series in 

 the table below,) was obtained during the month of June, when there 

 was no artificial heat in the building, and the temperature varied but 

 little during day and night. 



The globe used for holding the hydrogen — shown in Fig. 1 in about 

 one sixth of its actual dimensions — has an interior capacity of 49 G 1.5 



cubic centimeters, and weighs 570.5 

 grams. The cap with the connect- 

 ing tubes was ground into the neck, 

 and this joint, as also the stopcocks, 

 was so carefully made that there 

 was absolutely no leakage, and the 

 globe would hold a vacuum indefi- 

 nitely ; as was shown frequently by 

 its remaining banc on the balance 

 for weeks together when exhausted 

 without change of weight. The ap- 

 paratus was made by Einil Greiner, 

 of 79 Nassau Street, New York, 

 whose careful workmanship greatly 

 contributed to the success of our 

 investigation. The details of the 

 stopcock are shown at the sides ; 

 and it will be noticed that, besides 

 the direct way, there is a side way 

 through the plug of the stopcock 

 independent of the first, by which, 

 when the stopcock is closed, a connection is established with the base 

 of the cock, through which the gas may escape. 



Assume now that the interior of the globe has been exhausted, and 

 a gas current established through this side aperture from one of the 



Fig. 1. 



