A few things need to be cleared up.

First of all, do we have 20 men and 20 females, or 40 of each?

Next, we need to test or assume that the populations variances are equal.

Both are t statitsics.

We need to know if we should pool the smaple variances or use satterwaite's approximation.

Based on the data (which is cheating) I would assume that the pop st deviations are equal.

In that case use the formula below .... Unequal sample sizes, equal variance

from ....http://en.wikipedia.org/wiki/T-test

The second problem is clear.

It's a one sided test.

The first one I would guess is a two sided test.