Intelligence and Miscellaneous Articles. 67 



Depression of the Zero. — Person supposed that thermometers 

 heated for a long time would be rendered incapable of a depression 

 of the zero-point ; but that expectation has not been realized : the 

 depressions produced when the thermometer is heated after a long 

 rest cannot be put an end to. The values of these depressions have 

 been determined by several authors ; and it has been ascertained 

 that similar experiments give numbers identical to within o, 01 

 when the thermometers are heated to 100°; and the errors do not 

 exceed o, 04 for higher temperatures up to 300°. The only pre- 

 caution necessary is to follow one invariable method of observa- 

 tion : that which appears to me preferable to the rest is to let the 

 thermometer cool in the air and then immediately observe the zero. 

 Some observers, after an experiment at 100°, plunge the thermo- 

 meter into a bath heated to 50°, and afterwards into cold baths to 

 accelerate the cooling: equally constant results are obtained by 

 this means ; but the position of the zero is about o, 05 lower than 

 in the first case. If the thermometer be heated in a large bath 

 holding 20 litres, and the bath and thermometer be left to cool 

 together during twenty-four hours, the zero will be about 0°'15 

 higher than in the first case. If the zero-point be taken before the 

 experiment, or if the experimenter wait some time before observing 

 it, the positions will be higher, but will be more constant if he pro- 

 ceed methodically. 



Permanent Elevation of the Zero. — The observations above referred 

 to, as well as all measurements to be made with a mercurial thermo- 

 meter, are singularly impeded by the permanent elevation of the zero ; 

 and the alteration of the coefficient of expansion of the glass which 

 accompanies that phenomenon falsifies all measurements. This 

 movement of the glass-particles varies enormously in its extent, 

 according to the circumstances. Thus in a few hours at 430°, or 

 in a few days at 355°, the zero can be raised 17° or 26°, whilst the 

 writings of M. Libri and later a publication of M. Meucci's estab- 

 lish that, of some thermometers preserved at Florence for more 

 than two centuries, the position of the zero has not notably changed. 

 A fact of especial importance to us is that the permanent elevation 

 of the fixed points, produced at a high temperature, preserves the 

 thermometer from the effect of heat in this respect at lower tempe- 

 ratures. Some thermometers which were kept for eleven days 

 heated to 355°, and were afterwards constantly, during two years 

 and a half, submitted to experiments at all temperatures up to 

 326°, showed again, after being heated for half an hour to 355°, 

 the same position of the zero, wiihin o, l, as after the first heating 

 to 355°. 



Thermometers intended for our ordinary laboratory experiments 

 should be heated, before graduation and calibration, for a week or 

 ten days in boiling mercury. That is the only appropriate way to 

 obtain instruments that preserve the value of the degree fixed during 

 graduation ; and the errors of thermometers which have not under- 

 gone this treatment may amount to 4° for a length of 300°. 



When the thermometer is intended to give lower temperatures, 



