[BARNES] TEMPERATURE OF RIVER AVATER 39 



Wire-Bri'hje. 



TliLs was one which hud been constructed for Prof. Calk-ndar by the 

 Cambridge Scientific Instrument Co. after his own design, and esijcciall}' 

 made to read to thousandths of a degree Centigrade. 



It consists of a bridge-wire, 20 cm. long, with sliding contact piece 

 and resistance coils, which are convenient multiples of the bridge-wire. 

 The coils were all carefully measured and compensated by Prof. Callendar. 



The arrangement affords a resistance box with bridge- wire of infinite 

 length, unatïected by temperature variation, and having a vernier reading 

 to l/lOO of a millimeter. 



(Talranoiiicter. 



This was one of Ioav resistance constructed for Prof Callendar by the 

 Cambridge Scientitic Instrument Co., an-d especially designed by him for 

 thermOmetric work. 



Deflections were read by means of a telescope with micrometer scale. 



Battery. 



A Leclanché cell of the dry pattern served to supply the current, 

 and was steady to a tenth of a milliampere when not allowed to remain 

 on circuit more than half an hour at a time. 



Acruracy of the readings. 



Under suital)le laboratory conditions it was possible to measure to the 

 ten thousandth part of a degree. In the present case, owing to other 

 ditficulties, no attempt was made closer than one thousandth, A difference 

 of one degree Centigrade in the temperature of the thermometer stems 

 would produce such a change in the resistance of the platinum coils as 

 would be measured on the bridge-wire by a length of 20 cm. By arrang- 

 ing the sensitiveness of the galvanometer to read 50 scale divisions for 

 every centimeter each division would correspond to one thousandth Y>art of 

 a degree. Even under the most favourable conditions the measurement of 

 thousandths of a degree is involved with care, but on the present occasion 

 this operation was unusually difficult. A number of magnetic articles 

 were distributed both inside and outside of the shanty, while the mass of 

 iron contained in the heating ap2)aratus was enough, in itself, to disturb 

 the galvanometer. 



The movement of a mass of iron caused a variation in the zero reading 

 of the galvanometer that could only be rectified by adjusting the control 

 magnet, involving a redetermination of the sensitiveness. It will not be 

 possible, therefore, to give in tabular form the difterences measured. It 

 is possible to set limits to the variations and in a few cases to actually 

 give them to the ten thousandth part of a degree. 



Through the kindness of Mr. Kennedy some alterations were made 

 in the shanty, the chief one being the erection of a stand for the insiru- 



