The total range of temperature in the pendulum case was from 22 -75 to 25 -21 

 C., and the maximum change during observations was at the rate of 0-6 per hour, 

 the corresponding change in temperature of the air in the cave being of the order of 

 2 to 3 degrees. That the range' was not larger, is due to the fact that observations 

 were always made in cold, clear weather when the temperature outside varied from 

 about 30 C. to 40 C., so that the temperature could be kept reasonably constant 

 by leaving the door (formed of sacking) open during the period of observation. In 

 this series, the temperature shown by the thermometer in the dummy pendulum was 

 estimated by eye to hundredths of a degree. 



Barometer. The density correction is calculated from the reading of the station 

 barometer (No. 1157), corrected in the ordinary way and read about every hour. The 

 total range during the series was from 734-0 mm. to 752-1 mm. 



Hygrometer. Observations of the percentage humidity of the air were not taken, 

 the correction for the small amount of vapour being almost negligible. 



Flexure. Observations for the flexure correction were made as under : 



Nos. 5 and 7. No. 21. 



June 19th .. .. 20-6X10- 7 



July 3rd . . 23-1 



sec. 



4th 

 5th 

 6th 

 7th 



17-2X10- 7 sec. 



14-2 



15-1 



21-G 



Mean . . 1G 22 



Probable error 1 -Ox lO' 7 sec. 0-9xlQ- 7 sec. 



16x10"' sec. and 22xlO~ 7 sec. have been taken as the corrections to be applied 

 throughout. 



In Table XXI is given the full analysis of the pendulum results of series A and, 

 in Table XXII below, they are gathered together so as to show the behaviour of the 



individual pendulums. 



TABLE XXII. 



37 



c 3 



