i9n] THE SENSE OF COLD 291 
Held Service as usual. 
This afternoon went on ski around the bay and bark 
across. Little or no wind; sky clear, temperature -25 0 . 
It was wonderfully mild considering the temperature — this 
sounds paradoxical, but the sensation of cold does not 
conform to the thermometer — it is obviously dependent on 
the wind and less obviously on the humidity of the air and 
the ice crystals floating in it. I cannot very clearly 
account for this effect, but as a matter of fact I have 
certainly felt colder in still air at - io° than I did to-day 
when the thermometer was down to - 25 0 , other conditions 
apparently equal. 
The amazing circumstance is that by no means can 
we measure the humidity, or indeed the precipitation or 
evaporation. I have just been discussing with Simpson 
the insuperable difficulties that stand in the way of 
experiment in this direction, since cold air can only hold 
the smallest quantities of moisture, and saturation covers 
an extremely small range of temperature. 
Monday , May 29. — Another beautiful calm day. Went 
out both before and after the mid-day meal. This morning 
with Wilson and Bowers towards the thermometer off 
Inaccessible Island. On the way my companionable dog 
was heard barking and dimly seen — we went towards him 
and found that he was worrying a young sea leopard. 
This is the second found in the Strait this season. We 
had to secure it as a specimen, but it was sad to have to 
kill. The long lithe body of this seal makes it almost 
beautiful in comparison with our stout, bloated Weddclls. 
This poor beast turned swiftly from side to side as we 
U 2 
