Forwards...always forwards.

Yes, clockwise ;)

Yes ,from younger to older.

At the level of subatomic particles, many interactions are time-symmetric. That is, they are equally plausible going forward or going backward in time. Imagine two electrons that approach each other, deflect, and move apart again in a new direction. Now "play the movie backwards" and it still makes sense. There is no preferred direction for time's arrow. At the macroscopic level it gets more complicated. Two billiard balls will approach, collide, and move apart just like the electrons; this system is time-symmetric too. You can't tell if the movie is being shown forwards or backwards. But as soon as the pool player "breaks" (shoots a ball into the pack) the balls will scatter in such a way that, if you "play the movie backwards", it seems highly implausible that all those balls would come together at just the right instant to send the cueball back to the player. Something is going on! Time symmetry has been lost -- now you can definitely tell which way is forward in time. A key observation is that when the billiard balls are racked up for the break, the system is in an orderly state. After the break it is disordered. Another name for disorder is ENTROPY. The Second Law of Thermodynamics (http://en.wikipedia.org/wiki/Second_law_of_thermodynamics) states that the entropy of the universe tends to a maximum. It's easier to mix white marbles with black marbles by dumping them together than to sort the mixed marbles back to their original containers. It's easy to stir red and yellow liquids together to make orange liquid. Sorting the individual molecules back to their original containers is impossible. These systems, constrained by the 2nd Law of Thermodynamics, are not time symmetric because entropy increases only in one direction of time -- decreasing in the other direction. Some physicists believe that the 2nd Law of Thermodynamics is what establishes the arrow of time and defines what we mean by past and future. Time always flows in the direction of maximum entropy in the universe.

No, time is a scalar. It can be defined only on the basis of magnitude.