Here's the path on the torus parametrized by $(t,t/12)$ (i.e. twelve revolutions in one directions for each revolution in the other direction):

And here's the same path on a 3D embedding of the torus:

Now, if we flip the coordinates (by reflecting over the diagonal), the path looks like this when plotted on the unit square:

It is very obvious that this path is related in a simple way to the one shown before.
In contrast, the 3D representation of the second path looks very different from the one shown before:

It was not obvious to me, when I first saw them, that the two 3D paths shown above were isomorphic.
Q: Is it possible to deform one of the 3D representations to resemble the second one (in the way that the two flat representations resemble each other)?