APPARATUS FOR MAINTAINING CONSTANT TEMPERATURE. 897 



and the regulator set, to remove the thermometer altogether ; 

 the lateral tube may then serve, if necessary, for the intro- 

 duction of a small tube conducting gas or vapour to the 

 specimen under observation. 



The gas regulator may be described as a mercurial thermo- 

 meter with its tube open above, and with two side tubes leading 

 from it, one near the bulb, the other near the top (fig. 3). To 

 the lower side tube is cemented a steel collar [h), in which a 

 screw of the same metal accurately fits ; by working this screw 

 the mercury may be raised or lowered in the thermometer 

 tube. A fine steel tube (/), with a slit at its lower end, passes 

 down a certain way into the thermometer, being cemented 

 around its upper orifice. The gas is made to pass down this 

 fine steel tube, and then up between it and the walls of the 

 thermometer tube ; finally, it is conducted out by the upper 

 side tube, and by means of india-rubber tubing to the burner 

 below (as indicated by the arrows in figs. 2 and 3). 



To " set " the regulator, when the central chamber of the 

 stage has attained the desired temperature, all that is 

 necessary is to turn the steel screw until the mercury is 

 forced up to the slit in the steel tube ; the gas is now cut off, 

 except what can pass through the slit, and the flame is con- 

 sequently very small ; the temperature of the water in the reser- 

 voir consequently tends to be diminished, and the mercury 

 in the thermometer tube to fall, but the moment this com- 

 mences more of the slit becomes uncovered, more gas passes 

 through, the flame is increased, and the temperature re- 

 established. It is easy to understand that if the steel screw 

 below is withdrawn somewhat, the mercury will not rise up 

 to the slit, and will not therefore cut off the gas until the 

 temperature of the water has risen proportionately higher 

 than before. By screwing out or in every needful variation of 

 temperature of the water in the reservoir, and through this 

 of the stage, may, as before said, be obtained. 



It is easy to fill the closed circuit before described— con- 

 sisting of the reservoir, the stage, and the connecting tube of 

 india-rubber— with water (which should have been previously 

 boiled and allowed to cool), and once filled it will remain so, 

 provided the india rubber besecurely " wired '' over the metal 

 so as to exclude the possibility both of leakage and of the 

 admission of air. The india rubber will, of course, readily 

 adapt itself to the varying volume of the fluid, consequent on 

 the changes of temperature to which it is exposed. It is im- 

 portant to employ as wide india-rubber tubing as the metal 

 tubes will allow, so that no obstruction may be offered to 

 the free circulation of the water. 



