We have put forward and substantiated the hypothesis that complexes of activity, which are the main sources of geoeffective disturbances on the Sun, were developed, at least sometimes, during so-called grand minima of solar activity. This hypothesis is supported by the observation of 1) large sunspots by the naked eye and by telescopes during the said periods, 2) the white flare in 1705, and 3) aurora polaris, as well as by the pulse pattern of temporal variations in Wolf numbers with characteristic pulse duration from 6 to 13 months, typical for development of complexes of activity.