Hi Alex, if you can calculate the covariance between to variables the hard work is done.

For your m by n matrix the covariance matrix will be a symmtrical n by n matrix where each element is the a covariance between of of the variables.

Assuming your variables are in columns as x1, x2, x3, ... xn, then the first element in the covariance matrix S will be s_{11} which is the covariance between x1 and x1, then s_{12} is the covariance between x1 and x2 and so on.

How does that sound?