Ad 1)
The solution can be also on a line between two corners (inclusive).
For example you have the following LP problem:
$\texttt{max} \ \ z=4.8x+6.4y$
$0.6x+0.8y\leq 480$
$0.4x+0.2y\leq 220$
$x,y \geq 0$
You can solve the objective function and the first constraint for y.
$y=\frac{z}{6.4}-\color{blue}{\frac{3}{4}}x$
$y\leq 600-\color{blue}{\frac{3}{4}}x$
You see that both have the same slope. This is one necessary condition to have a multiple solution. This can be shown graphically. The grey area represents the permissible region. The green and the red line are the constaints. The blue lines represents the objective function, which is shifted towards the limit of the permissible region. At the end of this process there is a match of the red line and the blue line. The solution is between $(0/600)$ and $(400/300)$. 
Ad 2)
If you have no idea of linear programming, then it would be a good idea to go over all solutions in order to find the optimal solution. But there a more efficient algorithms, i.e. branch and bound. Here you solve first the LP-problem. Then you search a integer solution which is close as possible to the solution of the LP-problem.