322 DISCOVERY REPORTS 



leading and clearing beacons would simplify the business and make it less precarious. 

 Accordingly we selected sites for five beacons, making three pairs of leading marks 

 (including one chimney) for anchoring and turning the ship clear of danger. Sub- 

 stantial beacons were erected, fixed, and charted before we left. 



Bay of Isles (Chart 3) 



On completion of Prince Olaf Harbour it was decided to proceed to the more or less 

 detailed survey of the Bay of Isles, which though itself of no very immediate importance 

 navigationally, gives access to two moderately good anchorages, namely Rosita or 

 Allardyce Harbour and Elephant Beach Anchorage. The remaining bays are open to 

 much swell as well as being exposed to very strong wind. 



A trip was accordingly made to the Bay of Isles for carrying out a preliminary exami- 

 nation, but a thick fog descended and spoilt the operation. On the first suitable day, 

 November 28, we proceeded with the work. On this particular day the weather was 

 gloriously exhilarating, with clear blue sky and delightful air — as an occasional day can 

 be in South Georgia. Leaving fairly early in the morning, we soon covered the short 

 distance to Cape Wilson and then made a complete coastal reconnaissance, following 

 the shore closely right round to Cape Buller and keeping a careful look-out for a likely 

 camping ground. We concluded that there were only three possible places, each with 

 . several disadvantages. We deliberated for some time, and finally decided on a spot on 

 the south side of Camp Bay as being the least unpleasant ; it had some shelter from most 

 winds and a jutting promontory gave protection, as I thought, from swell. The beach 

 was stony and the anchorage for the 'Alert' was in the middle of kelp, but about the 

 right depth, 4 to 5 fathoms, with a bottom of stones and sand. The tents were erected 

 on the stones on the beach, mitigated by as much moss and tussac as we could find. 



The first four days proved quite fine, and notwithstanding the distances involved, a 

 great deal of the triangulation was carried out. The bay measures 9 J by 8 miles and 

 contains nine islands, nearly all of similar height but in size varying from Albatross 

 Island, approximately i mile square, to little Dot Island, a mere large rock. 



Then, on a rainy but calm day, it was decided to replenish fuel at Prince Olaf Har- 

 bour. This was always our main preoccupation, as we could carry a shorter supply of 

 that than of anything else. We obtained our fuel and accomplished the round journey in 

 less than five hours, bringing back another 50 gallons of petrol. The next day was 

 again fine and further work was carried out. 



The weather during the following night, however, became worse: a breeze from the 

 east began to bring in a swell, which rapidly became fairly heavy. Until this time our 

 practice had been to haul the pram on the beach at night. The beach here, however, 

 was so stony that even the pram began to suff'er, notwithstanding her extra bilge planks ; 

 accordingly she was anchored a short distance off" with a 30-lb. anchor. At about 4 a.m., 

 December 2 , we were awakened by the noise of her being thrown on the beach, whereupon 

 O'Connor and the men turned out and hauled her up, undamaged, but unfortunately 

 with one oar missing. This turned out to be a real misfortune, as with wind and swell still 



