It seems that SSL sites aren't crawled properly, because wget decides that the scheme is not to be followed. Offending code appears to be limited to only 3 lines located in recur.c: (version 1.8.1)
Line 440: change to if (u->scheme != SCHEME_HTTP && u->scheme!= SCHEME_HTTPS Line 449: change to if (u->scheme == SCHEME_HTTP || u->scheme == SCHEME_HTTPS) Line 537: change to if (opt.use_robots && (u->scheme == SCHEME_HTTP || u->scheme == SCHEME_HTTPS)) Thomas