Rowland's Thermometers with the Paris Standard. 11 



to be described presently, this box was kept open. As soon 

 as the water had attained to a uniform temperature, and 

 everything was ready to make a comparison, the stirring was 

 stopped, and the lids of the box were closed. The bulbs were 

 thus shut off completely from any convection currents, and 

 in this manner the thermometers could be kept at a constant 

 temperature For a considerable time. The water was stirred 

 by three paddle-wheels operated by means of a crank on the 

 outside of the box. 



In order to heat the water in the tank to bring it to any 

 desired temperature, a brass tube ran along it from end to 

 end near the bottom. This did not communicate with the 

 water of the tank, but opened into the air at each end. Steam 

 was passed into it until the w r ater reached the temperature 

 desired. 



Fastened to the lid of the tank were upright brass posts 

 which held the ways on which the micrometer telescope slid. 

 The telescope had freedom to move in two horizontal directions 

 at right angles to each other, so that it could be placed quickly 

 over any part of the tank, and used to read any thermometer 

 at any part of its scale. The thermometers rested on little 

 brass beams which were fastened to the brass rim of the tank. 

 As this rim was clamped securely to the brass lid which 

 carried the ways, the fourth requirement was secured. When 

 the tank was full of water and the lid on, the thermometer 

 bulbs were about 5*7 centim. below the surface of the water. 



d. Zero Tank. 



In taking the zero points of thermometers in the ordinarv 

 tank it is necessary to plunge the thermometer down into the 

 ice until the zero division is a good deal below the surface, so 

 as to make sure that the stem is all at the temperature of 0° C. 

 When a reading is made, the thermometer is raised until the 

 zero point is just visible. When this is done, not onlv is the 

 stem apt to become covered with water in films and drops, 

 which make it difficult to read the position of the mercury 

 thread, but it is also apt to become covered with dew when 

 the laboratory room, as is usual in this country, is at a 

 temperature of from 16° to 20° C. When the zero point is 

 near the bulb there is also, it seems probable, some danger 

 that a little warm water may trickle down the stem of the 

 thermometer and find its way to the bulb in sufficient quantity 

 to make a difference of several thousandths of a degree. In 

 order to do away with these difficulties it was decided to have 

 the stem immersed in the ice for some distance above the 

 zero point, and to read the thermometer through a tube which 



