A THEORY OF BLIZZARDS 
373 
as we have found, volumes of cold air at different tempera- 
tures are very loath to mix, the condition could not be 
relieved by any gradual process, but continues until the 
stream is released by some minor cause, when, the ball 
once started, a huge disturbance results. It seems to 
be generally held that warm air is passing polewards 
from the equator continuously at the high levels. It 
is this potentially warm air which, mixed by the disturb- 
ance with the cold air of the interior, gives to our winds 
so high a temperature. 
Such is this theory — like its predecessor it is put up 
for cockshies, and doubtless by our balloon work or by 
some other observations it will be upset or modified. 
Meanwhile it is well to keep one's mind alive with such 
problems, which mark the road of advance. 
Sunday, August 6. — Sunday with its usual routine. 
Hymn singing has become a point on which we begin to 
take some pride to ourselves. With our full attendance 
of singers we now get a grand volume of sound. 
The day started overcast. Chalky is an excellent 
adjective to describe the appearance of our outlook when 
the light is much diffused and shadows poor ; the scene 
is dull and flat. 
In the afternoon the sky cleared, the moon over 
Erebus gave a straw colour to the dissipating clouds. 
This evening the air is full of ice crystals and a stratus 
forms again. This alternation of clouded and clear skies 
has been the routine for some time now and is accompanied 
by an absence of wind which is delightfully novel. 
The blood of the Crozier Party, tested by Atkinson, 
