spaced. At times, however, it was found that the spacing was exceedingly irregular and, 

 in the extreme case, a flash might be so displaced as to appear to move backwards. 

 The cause of this is not easy to determine. That the low temperature in the cave was 

 not the sole cause of the trouble seemed clear from the fact that the coincidence 

 apparatus did not work quite satisfactorily even under the more favourable temperature 

 conditions of the second year, though undoubtedly the variations were of much 

 smaller magnitude in these later observations. The electrical connections appeared 

 to be quite good, and one is forced to the conclusion that irregularity in the action of 

 the coincidence spring or of the clock contacts was the cause of the anomalous results 

 noted above. 



For this reason, it was decided to devise a better form of apparatus which would 

 have no moving parts, and which would, therefore, not be subject to the same defects 

 as this apparatus. Unfortunately, there was no gear suitable for making such 

 a piece of apparatus, so one had to be content to do the best possible with the coinci- 

 dence apparatus, and merely to ensure that the necessary parts of the new design should 

 be waiting on the return to New Zealand and be available for subsequent work. 



Rates of chronometers. The clock used was the pendulum clock lent to the Expe- 

 dition known as " S.C." It was hung on the wall of the living-hut close to the chro- 

 nometer corner, and was daily compared, by the method of coincidences, with a mean time 

 box chronometer by Kullberg. This was denoted by the letter " E," and was itself 

 rated fortnightly by Commander Evans by sights on stars east and west.* Considerable 

 difficulty was experienced with the clock " S.C." as regards irregularity of rate, partly, 

 no doubt, due to the fact that the steel strip holding the pendulum had been broken 

 in transit from Potsdam to Christchurch, and had to be replaced by a strip made by a 

 local jeweller in Christchurch. Another serious difficulty was caused by a tendency 

 of the hut to shift on its foundations and thus throw the clock out of the vertical. 

 On one or two occasions, indeed, the hut shifted so much from the level that the 

 escapement was unable to act on one side of the escapement wheel, owing to the great 

 difference in the amplitudes of the pendulum on either side of the centre. 



The history of the clock in the winter of 1911 is as follows : On June 17th, the 

 clock was set up on the wall of the hut and was running smoothly by the 21st. It 

 went satisfactorily with a decreasing rate until July 15th, when the weights on the 

 pendulum were altered. From July 18th to 22nd, it ran in an unsatisfactory manner 

 and, on the 22nd, a new steel strip (a spare one made in New Zealand) was put in the 

 place of the old one. By July 23rd, it was again running smoothly, and kept a fairly 

 constant rate by comparison with " E " until it stopped on the 31st. On this date, 

 " the clock was found badly off the vertical with the driving weight touching the side 

 of the case and the escapement lever out of gear." 



The clock was rehung, and from August 1st to 3rd maintained a constant rate, but, 

 on the 3rd, was found stopped in the same manner as before. 



* Altitudes by sextant. 

 30 



