I concur in the obervation. Really, though, the arrow is wrong in
being too ambiguous -- it should show the passage, completely,
not merely point to the entrance (and then it would be clearly
wrong or right).
Fig.1 is bad re the apparent over/under of line,
and I will emphasize that in fact the apparent UNDER of the
end SHOULD BE what is done, bringing the hand under and
around-through, so that at Fig.3 the loop cast into the S.Part
doesn't require magic (or copious hair gel) to hold it in place
-- the S.Part would lie atop its continuation out of the loop,
not below it! This reversed orientation is every bit as easily
done, and results in presenting the face of the bowline that
is better viewed; I think that much of the confusion and difficulty
with the knot stems from viewing it from the opposite face (Fig.5).
- - - - - -
Here is the seed of an interesting case of plagiarism: in his final
comment about the Dbl.Bwl, Geoffrey claims that it has a breaking
strength of 70-75%, making it stronger than the single (YMMV on
this -- I've seen quite contrary figures, putting them as equal).
In a Maria Constantino [sp?] book she borrowed this snippet
as ". . . is 70% stronger than the single bowline",
which given 65% for that renders the double > 100% !!
-- and so it goes . . .