8 Dr. W. S. Day on a Comparison of 



b. Micrometer Telescope. 



The thermometers in the comparison-tank and in the zero 

 tank were always read by means of a micrometer telescope. 

 The micrometer was not an eyepiece micrometer, but there 

 were cross-hairs in the field of view, and the whole telescope 

 was moved parallel to itself by the micrometer motion. The 

 telescope was so constructed that it could be used to read the 

 horizontal thermometers in the comparison-tank, or the 

 vertical thermometers in the zero tank. In reading vertical 

 thermometers the telescope was fastened to an upright frame, 

 consisting of two vertical steel bars. It then became in effect 

 a cathetometer. In all cases care was taken by means of 

 levels and plumb-lines to have the axis of the telescope per- 

 pendicular to the thermometers. The telescope magnified 

 about ten or eleven times. 



c. Comparison- Tank. 



The comparison-tank was devised to fulfil the following 

 requirements : — 



1. The thermometers must be easily removed for the pur- 

 pose of taking zeros. 



2. They must remain at a practically constant temperature 

 for four or five minutes. 



3. They must be easily and quickly read by means of a 

 micrometer telescope sliding on ways above the tank. 



4. The supports on which the thermometers rest must be 

 so solid and firmly fastened to the ways on which the telescope 

 slides that there will be no possibility of any accidental relative 

 motion of the telescope and the thermometer. 



To fulfil the first requirement it was decided to have a 

 horizontal tank, as explained farther back. The tank was 

 made of sheet copper in the form of a rectangular box 

 92 centim. long, 21 centim. wide, and 23 centim. deep. A 

 section of it is represented in fig. 1, and a cross-section in 

 fig. 2. The top consisted of a movable lid made of a brass 

 casting. This lid had a flange that was planed on the under 

 side so as to fit a planed brass rim forming the top of the 

 sides of the tank. A piece of sheet indiarubber formed a 

 packing between the two parts, to prevent the escape of water. 

 After the lid had been placed in position, it could be screwed 

 down tightly by means of a number of C-clamps placed 

 around the sides. The lid was made in the form of a frame 

 holding a piece of plate-glass, which was a window through 

 which the thermometers could be read. This window was 

 about 1*5 centim. from the stems of the thermometers. 



