Plotting the off-diagonal value of a covariance matrix

Question

This block of code outputs all 10 of my covariance matrices and plots every point in the 2x2 matrix.

for i in range(10):
    columns = datawithoutmean[:, i*2:i*2 + 2]
    cov = numpy.cov(columns.T)
    print(cov)
    matplotlib.pyplot.scatter(cov[:,0], cov[:,1], c = 'r', marker = '.')
matplotlib.pyplot.show()

How can I specifically output the off-diagonal value for each co variance matrix and plot it using pyplot?

For reference the first co variance matrix looks like this:

[[10.34020531  -0.01203439]
[-0.01203439    2.06085007]]

I want to plot the off-diagonal (the co variance between the two columns), so in this case it would be -0.01203439.

Edit: I found out I can get the off-diagonal like this: Ok so If I do this, it'll output the co-variance value between the two columns:

for i in range(10):
    columns = datawithoutmean[:, i*2:i*2 + 2]
    cov = numpy.cov(columns.T)
    off_diagonal = cov[0][1] # covariance value

But how can I plot those values in a scatterplot using pyplot?