' A 



APPENDIX 175 



temperatures, and 100 C. ; but this does uot diminish their value 

 for general use. The exact reading of a thermometer may easily be 

 obtained by sending it for correction to Kew. In using a thermo- 

 meter ifc is of great importance to read accurately. To do this 

 practice is necessary. This caution is especially needful in comparing 

 temperature-changes in different kinds of matter, where a large 

 quantity of matter, say fifty grams, may be changed in temperature. 

 A mistake of half a degree is frequent in a beginner. This would 

 mean a mistake of twenty-five times the unit temperature-change 

 (or, as it is called, twenty-five units of heat). At the same time it 

 must be remembered that the temperature of the hand may easily 

 give rise to an incorrect reading. Errors of this kind must be cor- 

 rected by taking as many observations as possible in as varied a 

 manner as possible. Mistakes in the numerical results of temperature- 

 changes are much more likely to be due to incorrect reading of the 

 thermometers than to inaccuracy of weighing. Accurate results 

 must not be expected, however, unless special precautions are taken, 

 either to allow for the temperature-changes which simultaneously 

 take place in surrounding bodies, the vessels, air, &c., or else to 

 reduce these to a minimum by non-conducting material. Change of 

 temperature taking place more rapidly the greater the difference 

 of temperature in the bodies undergoing it, it is of course advisable 

 to adjust the quantities of matter under observation in a given ex- 

 periment, so that mutual changes may produce no very great differ? 

 ence of temperature between them and external bodies, such as the 

 air. This is not always possible. 



In order to measure the expansion of solids, paper scales pasted 

 to wooden rods or wooden scales to which sliding verniers are attached, 

 will be found sufficiently accurate. Inaccuracy of result is more 

 likely to lie in incorrectness of reading than in incorrectness of 

 graduation, and for this reason the vernier should carry an eyepiece 

 fitted with cross-wires. These may be obtained from a wholesale 

 maker very cheaply. 



4. Observations of Fall of Bodies to the Earth. It is convenient, 

 in observing the fall of bodies to the earth, to have a pulley fixed as 

 high as possible in a room, and, passing over it, a cord, by means of 

 which the electro-magnet, together with the wires needed, may be 

 hauled up to a considerable height with the bodies for experiment 

 adhering. A key for breaking contact will be required. Direct 

 contact of the body with the magnet may be prevented by means of 

 a brass guard more neatly than by a piece of paper. 



Results with ' Atwood's ' machine are not likely to be more than 



