embryo hut. On the 23rd, it had been restored to its former state and was demolished 

 again by wind on the 27th. On this occasion, a heavy 80-lb. wooden beam nailed up 

 as a ridge pole was torn off and hurled a full 40 feet. On the 28th, the hut was finished 

 and the canvas roof put on, the whole being well banked with snow. Succeeding blizzards 

 were unable to move the hut. It was, however, reluctantly abandoned after several 

 days' trial, as it was found impossible to keep the hut at a workable temperature, or 

 even to keep it free from drift snow. 



The only alternative was to place the pendulum apparatus in the living hut itself. 

 This had the great advantage that the temperatures were above freezing-point, and 

 that there would be no danger of short circuits or breaks in the wires from the clock 

 to the coincidence apparatus. In the preceding series, this had caused trouble owing 

 to the difficulty of locating defective wires tinder 5 feet of snow. The only objection 

 connected with the use of the pendulums in the hut seemed to be the trouble of finding 

 a rigid base on which to place the pendulum stand. Owing to our reduced numbers 

 now, the hut was quite large enough to accommodate the pendulum instruments, and 

 Debenham kindly lent the use of his photographic dark-room for the work. This 

 was of peculiar advantage, in that the dark-room formed a separate compartment 

 whose temperature could be kept fairly constant by closing the door giving access to 

 the living-room. 



In the floor of this dark-room, a large hole about 2 feet square was cut, and the 

 frozen volcanic sand attacked with picks until an excavation about a foot deep was 

 made in the frozen mixture. A large kenyte boulder just able to enter the hole in the 

 floor was next sledged into the hut and placed in position in the hole prepared for it. 

 The top surface of this was prepared roughly flat and was level with the floor of the 

 hut. Around the boulder was next placed a quantity of the volcanic sand and a pail 

 of water poured over it and allowed to freeze. At intervals during three of four days, 

 sand and water were added (as fast as the water could be melted on the cooking range), 

 until the boulder was firmly cemented to the ground and to about a cubic yard of the 

 volcanic gravel. 



Conical holes were next drilled into the surface of the boulder for the reception of 

 the levelling screws of the stand, and the hole in the floor covered with a loose double 

 layer of Burberry cloth. This was designed to keep out draughts from below, without 

 making rigid connection between the boulder and the unstable floor of the hut. Small 

 holes in the wind-proof covering permitted the levelling screws to rest directly on the 

 boulder. The next step was to make a more rigid support for the pendulum clock, and 

 another hole was made in the floor, and a heavy solid pillar of Jarrah wood, 1 foot wide 

 and 6 inches thick, was cemented into the ground with water and gravel. A heavy 

 screw supported the clock. In this position, the clock hung for several months without 

 stopping (being now independent of changes in level of the hut). 



That the temperature of the pendulums might remain more constant, a small square 

 was cut from one of the walls of the dark-room, and the flashes from the coincidence 

 apparatus observed through this hole in the wall. Through the same hole, the face of 



46 



