Those four variations of the H bend ( Hoban s bend, or Hitch bend ) come in two pairs. Let us call them A1, A2 and B1, B2. Each member of each pair is topologically equivalent to the other, and can be transformed into the other by a simple re-arrangement of the tails into the knot s nub. To indicate this fact, I decided to change the provisional labels I have used, in this way : A ( previous)=> A1 ( present ), D => A2, B => B1, C => B2. See the attached pictures, for the new labels. At the "1" pair, because of the particular location of the two tails into the knot s nub, the tail of each link is pushed against the tail of the other link. So, at the "1" pair, the "white" tail of the less convoluted hitch component ( the "white" rope ) is "first" pulled by its own standing part, and "then" it is pushed onto the "orange" tail - and, finally, it is secured by the friction forces generated along its mutual contact area with this "orange" tail -, while at the "2" pair the "orange" tail is the one which is "first" pulled by the "white" standing part, and "then" it pushes the "white" tail ( which, finally, by this sequence of pulling and pushing, it is squeezed in between the two legs of the "orange" standing part s first curve, on the one hand, and the straight segment of its own "white" standing part, on the other). The way the tails are secured resembles the Double Harness bend, in the first case, and the Angler s loop, at the second.