Effect of Pressure on Thermometer-bulbs. 411 



mometers under certain circumstances still remains to be 

 discovered*. 



I may mention one source of error in thermometric work 

 which attains considerable dimensions when dealing with tubes 

 of very fine bore. These tubes, even when of the most perfect 

 description obtainable, generally possess a few points at which 

 the mercury column experiences a difficulty in passing : the 

 mercury, when it has reached such a point, sticks there an 

 appreciable time, and then passes it suddenly with a jerk ; 

 sometimes, even, the mercury sticks so persistently that the 

 column will separate sooner than pass it. These points do not 

 indicate any contraction which is sufficient to affect the results 

 of calibration, and are probably due to some difference in the 

 nature of the glass, for they may be developed by heating 

 the thermometer- tube externally with a very small flame up 

 to about 400°. The error of taking a reading while the 

 mercury is sticking at such a point may, I estimate, amount 

 sometimes to as much as 05 millim. All delicate thermo- 

 meters should be carefully examined in order to ascertain the 

 position of such points, and they should be avoided, if possible, 

 in any work with the instrument. 



In any thermochemical work in which the effect of tempe- 

 rature on a given reaction is being studied, many of the 

 sources of error inherent in the use of thermometers may be 

 avoided by using the same portion of the stem of the instru- 

 ment, whatever the actual temperature may be. To effect 

 this, the zero-point is altered in each experiment by removing 

 some of the mercury into the upper chamber of the thermo- 

 meter. Formerly I removed the requisite amount of mercury 

 by the application of a very small flame to a point just below 

 this chamber ; but I now adopt a method which is much 

 safer, more expeditious, and equally exact. A fine tube, 

 somewhat wider than the stem of the thermometer, is affixed 

 to the upper end of the stem, and in this tube there is a small 

 contraction or " knife-edge,'"' sufficiently wide to permit of the 

 mercury passing it either upwards or downwards, but yet so 

 narrow that a slight swing of the instrument will cause the 

 column of mercury to break off at it, that portion of the 

 mercury which is above the knife-edge passing up into the 

 upper chamber. By this means any point on the stem may 

 be adjusted with ease to within o, 02 of any given tem- 

 perature. 



* For an unexplained instance of non-accordance of results olbtained 

 with different thermometers, and also with different portions of the stem 

 of the same thermometer, see Chem. Soc. Trans. 1887, pp. 304, 322. 



