ANOMALOUS BEHAVIOUR OF DELICATE BALANCES, ETC. 



403 



Very similar results were likewise obtained with the other balances. For example, 

 the respective R.Ps. of the Oertling long beam and a Sartorius triangular beam were 

 found to undergo, with variations in temperature and load, the approximate displace- 

 ments set forth in Table IV. 



TABLE IV. 



So far as we are able to judge, all balances are liable to be affected in this wholly 

 unexpected manner. 



Of tJie Variations in the Temperature within t/te Balance Case, as Observed with a 



Bolometer. 



Of the numerous precautions which must be observed when it is desired to attain 

 to a high degree of accuracy in weighing, possibly none are more important than those 

 which tend to ensure uniformity of temperature throughout the beam. 



We have already shown that any deviation from strict equality in the temperature 

 of the two arms must, in such cases, be considerably less than 1/100 C. Now it 

 appeared probable that small temperature differences would generally exist within a 

 balance case, and therefore within the arms, even in spite of all the usual precautions. 

 We therefore decided to investigate this matter somewhat minutely and endeavour to 

 discover both their nature and magnitude under the ordinary conditions. 



For this purpose a mercury-in-glass thermometer, even if constructed so as to be 

 capable of indicating very small differences in temperature, would be wholly unsuit- 

 able on account of its sluggish nature. Hence it was decided to employ a modified 

 form of the Langley bolometer. This was constructed in the following manner : 



Two thin rectangular sheets of mica, M, M', each 10 cm. long and 2 cm. wide, were 

 supported, as shown in Fig. VII., by ebonite rods, E, E, the distance between them 

 being about 20 cm. Upon each sheet was then wound, in spiral form, about one metre 

 of platinum wire having a diameter of 1/20 mm. ; the ends of the platinum wires 

 were soldered to the copper wires r, s, r', s', these latter were similarly joined to the 

 copper straps a, b, d, and c. Very fine glass capillary tubes, t, t, were thrust down 

 on either side between the coils of platinum wire and mica. The spirals were thus 

 made, as far as circumstances permitted, air lines. 



3 F 2 



