I don't think the results are going to be very interesting but you could do what you are asking as follows:
perp = 365;
perd = 1896.59;
np = 2*pi / perp;
nd = 2*pi / perd;
dT = 2*pi / (np - nd);
T = dT:dT:dT*ceil(100*perp/dT);
This is just going to add a constant increment to theta every dT. If you plot(T, theta) you will get a boring straight line. You could, of course, apply a mod by 2*pi but this won't be much more useful (IMHO).
Anyway, I hope this answers your question. If not, please clarify.