Meteorological and Astronomical Notices. 179 



when it was found that, in addition to expansion from heat, which 

 could be measured, there was other expansion from moisture, 

 which could not be measured accurately, and other alteration still 

 from some sort of working or reaction of the late organic forces, 

 which had vivified the wood, and made the tree, and are more difficult 

 than ever to examine into, and seem to defy all our means of mea- 

 suring their quantity and computing their effects. While bone or 

 ivory, anything in fact which has had life in it, is as little to be 

 trusted as wood. 



Recourse was then had to metals, and men revelled in the fancied 

 simplicity and certainty of the corrections to be applied to deduce the 

 true length of a bar at any temperature. Bars of all metals, and in the 

 state of metal, cast or wrought, hammered or rolled, round, flat, 

 thick, thin, hollow, and solid, were all indifferently used : the ther- 

 motic expansion being ascertained in each case, by trying the length 

 first in a freezing mixture, then in boiling water, and taking yl^j^ of 

 the difference so found, as the correction due to 1° Fahr. 



But now it is found by close microscopic measurements, that ac- 

 cording to the degree to which the particles of the bar were dis- 

 tressed in the metallurgic operations of rolling, hammering, &c., so 

 will it have a greater tendency to return, though slowly and through 

 long intervals, to some former shape. And as we cannot ascertain 

 all the blows and severe trials which the bar may have undergone 

 in its process of formation, or probably correct for them, if we did, 

 for the effects will vary with every different metal and every variety 

 of each metal, we can only eschew all recently manufactured bars, 

 and prefer cast to wrought metal. 



Next it has been ascertained that no metal expands uniformly 

 with the mercurial thermometer, and therefore the making the cor- 

 rection equal to so many ^^-^ parts of the whole expansion, as the 

 thermometer should be above 32° F., is by no means true for points 

 between 32° and 212°. 



And lastly if the bar be heated up to 212°, it will be found not to 

 come down again at once on cooling to its former length at 32°. As 

 therefore so high a temperature is unnecessary, because we never 

 have to use the rods, or to make measurements with them in an atmos- 

 phere of that temperature, the once favourite plan of boiling or roast- 

 ing bars, so as to get their expansion indicated to a great extent, is 

 now given up ; and the better method is adopted of leaving the 

 bars altogether to themselves and to nature. They are allowed to 

 take any temperature that they please under the usual changes of 

 climate ; their lengths are carefully watched under all those slightly 

 varying alterations of natural heat ; observations are greatly multi- 

 plied ; and equations of condition at last bring out the true law of 

 expansion. 



M 2 



