Observations luitJi Rater's Pendnhivis. 5 



The temperature was determined as in Melbourne by means 

 of the thermometers, K667 used in the inverted position, and 

 K668 in the erect position, both attached to the dummy ; and 

 as a check on any irregular variations of temperature, the two 

 Richard thermographs were employed. The tracings furnished 

 by these were in all cases so regular, that the mean of the 

 thermometer readings, with the correction - 0° 13 applied, could 

 be always taken as representing the mean temperature of the 

 pendulum with sufficient accuracy. The thermometers and 

 barometer were read before and after the observations at the 

 beginning and end of a set of swings ; and each recorded reading 

 is therefore the mean of four observations. The fluctuations in 

 barometric pressure were also observed by means of the Observa- 

 tory barograph ; though not large enough to sensibly effect the 

 pressure correction, they influence the observations in another 

 manner, as described in the concluding paragraph of this paper. 



Arc of Vibration. 



This was read on the arc scale behind the tail-piece of the 

 pendulum, as in previous observations with this apparatus. 



Levelling. 



The agate planes of suspension for pendulums No. 4 and No. 6 

 were adjusted to horizontality with the aid of two small but very 

 sensitive levels sent out with the apparatus ; each of these stands 

 on three sharp points. The agates belonging to No. 11, being 

 cylinders instead of planes, could not be adjusted with these 

 levels ; the two flat-based levels sent out with this pendulum are 

 very sluggish, and not very sensitive ; I accordingly employed a 

 very delicate flat-based level, kindly lent me by Mr. Ellery. The 

 planes generally remained in good adjustment as tested by 

 relevelling at the close of the series for each pendulum. 



Observation of Coincidences. 



In setting up the apparatus the observing telescope had to be 

 rotated to the left of the vertical, so that the observer sat with 

 the pendulum stand on his left. The disappearance and reappear- 

 ance of the apparent left edge of the image of the disc on the 



