408 ME. J. J. MANLEY: OBSERVATIONS ON THE 



Accordingly a small chamber, represented in Fig. X., was built up round tbe beam,* 

 within the maker's case. The base-plate, c, d, e, f, was of aluminium, having a 

 thickness of 2 -2 mm., suitably slotted and perforated for the passage of the pointer 

 and hooks by which the pans are suspended. The ends, E, F, were made of well- 

 seasoned mahogany 11 mm. thick, whilst the front, back, and also the top of the 

 chamber were of 8 mm. thick plate-glass. The front glass shutter was removable 

 after dropping the holders, /;, k. Discs r, s, t of thin aluminium were fastened to 

 the pointer and pan hooks a little below the base-plate. 



That this addition to the equipment of a delicate balance not only enhances the 

 reliability of measurements made with the instrument, but also enables the observer 

 to obtain his results with a minimum expenditure of time, is, we believe, shown by 

 the following evidence. The bolometer was introduced into the inner or beam case, 

 which was then closed. After a considerable interval the following experiments were 

 made : 



Experiment I. The ordinary sliding door of the balance case was opened and 

 closed a number of times without producing the slightest effect upon the bolometer. 



Experiment II. Nine weights were introduced and then removed in the usual 

 manner six times in succession. The bolometer was quite unaffected. 



Experiment III. The observer's closed hand was held close to the II pan for 

 5 minutes. During the first 3 minutes no visible effect was produced ; but during 

 the remaining 2 minutes the measured temperature difference within the beam case 

 was equal to +1/1000 of a degree C. 



Experiment IV. This was similar to the preceding experiment, but the observer's 

 hand was held close to the L pan. There was no observable change during the first 

 3 minutes, but at the expiration of the remaining 2 minutes the temperature 

 difference amounted to slightly more than -1/500 U 



Experiment V. When the experimenter's hand was held close to the pillar and 

 between the pans, a temperature difference amounting to 1/5000 of a degree C. was 

 set up at the end of 5 minutes. 



Experiment VI. A sensitive mercury -in-glass thermometer was placed close 

 beneath the aluminium base-plate and Experiment I. repeated. During 5 minutes 

 the thermometer rose l - 9 C. 



The results obtained from Expemment II., together with others found by repeating 

 the observations in the absence of the beam case, are given in the form of the curves 

 seen in Fig. XI. 



These experiments show that the inner chamber is highly effective, and that its 

 presence ensures a practically perfect uniformity of temperature throughout the entire 

 beam. 



Under ordinary conditions, and in the absence of some such device as the above, 



The balance chosen for this experiment was one by GALLENKAMP. Either of the other balances 

 would have been equally convenient. 



