I have read that statistical analysis of the distribution of the known non-trivial zeros of the Riemann zeta function has given a fractal dimension of about 1.9. Nice, but is there anything in any of the (non-statistical) equations that make the self-similarity evident?