318 ANNUAL REPORT SMITHSONIAN INSTITUTION, 1937 



lay all day in our sleeping-bags with drift and gale reaching 50 miles 

 per hour. By November 30 there were huge drifts around the plane, 

 and the cockpits were full of snow. We were unable to get into com- 

 munication with the Wyatt Earj), although on November 30 we got 

 three time ticks from Buenos Aires. We were 600 miles from Little 

 America and probably had not enough fuel left to get there. I con- 

 sider this stay at Camp III as the low-water mark of our flight. 



However, with our prospect appearing so dark, our situation was 

 improving. At Camp III we lucidly thought of the simple expedient 

 of adjusting the bubble of our sextant on the snow horizon when the 

 index read zero. Tliis showed that the sextant had after the first 

 observation developed an index error of 82 minutes of arc, and that 

 it had been apparently constant. We reduced it to an uncertainty 

 of about 4', enabling us to fix our position and to set a direct course 

 to Little America. Once the sextant was put in approximate adjust- 

 ment, our navigation problem became a simple one. AU observa- 

 tions, except the very first one, were corrected for the determined 

 index error of 82' and the positions reworked with impressive results. 



The second observation, taken in the air at 18.54 on November 23, 

 showed that the plane was behind schedule. The big discrepancy in 

 the estimated air speed of 145 miles per hour and the actual ground 

 speed is accounted for by several factors: (1) The substitution of 

 skis for wheels causing an unexpectedly heavy drag; (2) a slightly 

 crumpled fuselage wliich altered the streamhne and thereby reduced 

 the speed; (3) unexpectedly heavy head winds; (4) low temperatures 

 which reduced the engine-power output; and (5) throttling down the 

 engine to save fuel. During the whole midsection of our flight, 

 from the time we left Eternity Range until we started on the down- 

 grade to the Ross Barrier, the prevailing wind blew from the oast 

 and southeast. Only twice did we have a north Avind, and then only 

 for a few minutes. We never had a west wind. But these factors 

 hardly account for the reduction of more than 25 percent in the speed 

 of the Polar Star. The measured speed at the beginning of the flight 

 when the plane was heavily laden, and the known speed of the plane 

 on the last two legs, was relatively much higher, so that an extremely 

 low ground speed of about 92 miles per hour was made on the first 

 and most dangerous leg of the flight. 



On December 1 we spent the whole day clearing snowdrifts from 

 the plane, which was one solid block of snow. To crawl in among 

 the controls with a teacup and clear away dry snow as fine as flour 

 was the worst job of all. On December 3, we tried to start up the 

 machine, but the magneto burned out. It looked as though we were 

 650 miles from the Bay of Whales with no hope of getting there. 

 When the blizzard abated we were able to cut snowblocks to erect 

 a shelter to the windward of our tent. For 8 days, until December 



