To investigate the location of the actual limit for the different behavior, the active regions with fluxes greater than 10 to the 20 Mx are investigated. After surveying the differences between large and small active regions reported in the literature, the birthplaces of the active regions are analyzed statistically. The analysis indicates that the birthplaces of the small active regions are almost always necessarily zones of remnant magnetic flux. This, it is noted, is in contrast with the larger ones. The limit between the large and small active regions is seen as approximately 10 to the 21st Mx. It is argued that if some magnetic flux of the decaying regions is being sunk down below the photospheric level (as suggested by numerical calculations and some observations), then some kind of re-apparition is to be expected at the solar surface. A tentative picture is proposed, whereby the small active regions, and perhaps the ephemeral ones as well, are the secondary effects after the decay of the large active regions and complexes of activity.