n s-shaped curve to a level about twenty or twenty-five feet higher
than that of the road along which I had been driving. The bridge had a
rail on its west side; but the other rail had been broken down in some
accident and had never been replaced. I mention this trifle because it
became important in an incident during the last drive which I am going
to describe.

On we went. We passed the school of which I did not see much except the
flagpole. And then we came to the crossroads where the trail bent west
into the town. If I had known the road more thoroughly, I should have
turned there, too. It would have added another two miles to my already
overlong trip, but I invariably did it later on. Firstly, the horses
will rest up much more completely when put into a stable for feeding.
And secondly, there always radiate from a town fairly well beaten
trails. It is a mistake to cut across from one such trail to another.
The straight road, though much shorter, is apt to be entirely
untravelled, and to break trail after a heavy snowstorm is about as hard
a task as any that you can put your team up against. I had the road;
there was no mistaking it; it ran along between trees and fences which
were plainly visible; but there were ditches and brush buried under the
snow which covered the grade to a depth of maybe three feet, and every
bit of these drifts was of that treacherous character that I have
described.

If you look at some small drift piled up, maybe, against the glass pane
of a storm window, you can plainly see how the snow, even in such
a miniature pile, preserves the stratified appearance which is the
consequence of its being laid down in layers of varying density. Now
after it has been lying for some time, it will form a crust on top which
is sometimes the effect of wind pressure and sometimes--under favourable
conditions--of superficial glaciation. A similar condensation takes
place at the bottom as the result of the work of gravity: a harder core
will form. Between the two there is layer upon layer of comparatively
softer snow. In these softer layers the differences which are due to the
stratified precipitation still remain. And frequently they will make the
going particularly uncertain; for a horse will break through in stages
only. He thinks that he has reached the carrying stratum, gets ready to
take his next step--thereby throwing his whole weight on two or at best
three feet--and just when he is off his balance, t