MESSRS. R. THRELFALL AND J. A. POLLOCK 



natural change which takes place every evening, the second rise of temperature 

 being due to the room being heated by the lights used for observation. These 

 observations gave us what we call a natural minimum. These examples are sufficient 

 to explain the reason of our ordinary procedure which is described in the next 

 paragraph. 



Method of Observing. 



In the afternoon, when the temperature of the room just commences to fall, a 

 watch is kept by one observer on the temperature of the box. When the rise of 

 temperature of the box becomes very slow, the other observer prepares to take 

 observations of the lever end. The preparation consists in carefully levelling the 

 instrument along a line parallel to the thread, placing the long striding level on the 

 microscope, and levelling cross ways until the sum of the readings of the two ends of 

 the bubble of this level is within a few tenths of a division of the sum found when 

 the level was last reversed. When the observer watching the temperature considers 

 that the temperature of the box will reach its maximum value within the next few 

 minutes, the lever is unarrested, a note being generally made of the time at which 

 this is done. 



A careful watch being kept on the levels, the image of the end of the lever is 

 taken slightly above the cross wire by screwing the tangent screw of the vernier arm, 

 the motion of the screw is then reversed, and the edge of the image brought down 

 to coincidence with the cross wire. (There is a difference in the reading of a 

 coincidence, depending on whether the image of the lever end is brought up to the 

 wire or down to it, of from 30" to 40" (sextant), so that settings are habitually made 

 downwards.) An observation of temperature is taken when the coincidence is exact. 

 Immediately a setting is made, the positions of the ends of the bubble of the microscope 

 level are noted, and the time, temperature and level readings entered. A reading of 

 the position of the vernier arm is then made at leisure. A complete observation 

 takes from three to four minutes. On levelling again, if necessary, it is seen that 

 the image of the end of the lever has moved upwards, so the former procedure is 

 repeated, until the end of the lever remains steadily in coincidence with the cross 

 wire. An independent entry is made in the note-book of the time at which the 

 temperature commences to fall. The image of the lever end is watched, levelling at 

 intervals, until it begins to come down, to ensure that the maximum reading has been 

 obtained. The lever is now arrested, and the vernier put at the constant reading 

 of 85 (as it happens), so that the amount of twist in the thread may be kept 

 constant with the exception of the time, never longer than an hour, occupied in 

 an observation. The aneroid and air thermometer readings are entered, the level 

 reversed and then removed from the microscope, and the observation is complete. 



