206 



READINGS OF THERMOMETERS. [January, 



On the 17th December I had noted that the mercury 

 thermometers did not cease to act at 40 '3, 40 '0, 

 40'0: this excited my attention; weekly I transcribe 

 the register. On the 28th I noticed that the registry of 

 all the mercurial thermometers ceased at the same indi- 

 cation, on the 4th January at 42 '3, by standard. Un- 

 fortunately I did not notice it earlier, but on the 23rd 

 of January, on my copying the rough register, I inserted 

 the following order : " The mark / to be prefixed to D 

 when the mercury is congealed ; p f, when partially con- 

 gealed. The numbers 19, 35, 34, 2, read to 40 and 

 below. The divisions below must be estimated, i. e, the 

 point where it congeals, or declines moving further." 

 He (Mercury) had stopped most religiously at 39 '3, 

 39-2, 39-0, 39-0; but in this case he was under 

 martial law, and he knew he must move. I suppose he 

 made his mind up to it, for I find the following curious 

 register, taking care to satisfy myself frequently by per- 

 sonal inspection. 



How ciime this change? 



