igii] 
THE ROOF GOES 
49 
We could not understand quite how the tent had been 
blown away, for we had taken extra precautions in setting 
it, and had got as nearly perfect a spread as possible. 
Moreover, it was in the lee of the hut, and we had buried 
the valance not only with heaps of snow, but with 4 
or 5 rocks on the snow in each bay, and to make things 
quite secure, the last thing before turning in Bowers and 
I had hoisted the heavy canvas tank, full of gear, almost 
more than one could lift alone, on to the weather skirt. 
We could only think that the same sucking action 
which lifted our roof also lifted the tent, or that it was 
twisted off its legs by getting caught sideways by a squall 
which came partly round the end of the hut corner. 
Anyhow, as it was gone, we decided to take the earliest 
opportunity of any light to go and look for it. 
Other things happened before this opportunity arrived. 
Sunday^ July 23, 191 1. — Bowers estimated the wind 
at force 11 and noted it as blowing with almost con- 
tinuous storm force, with very slight lulls followed by 
squalls of great violence. 
About noon the canvas roof of the hut was carried 
away, and the storm continued unabated all day, but 
latterly without much drift. 
It happened that this was my birthday — and we 
spent it lying in our bags without a roof or a meal, wishing 
the wind would drop, while the snow drifted over us. 
The roof went as follows. We saw, as soon as light 
showed through the canvas in the early morning, that 
the snow blocks on the top had all been blown off, and 
that the upward strain was now as bad as ever, with a 
VOL. II. E 
