A time-latitude diagram where spotgroups are given proportional relevance to their area is presented. The diagram reveals that the spotted area distribution is higly dishomogeneous, most of it being concentrated in few, small portions (“knots”) of the Butterfly Diagram; because of this structure, the BD may be properly described as a cluster of knots. The description, assuming that spots scatter around the “spot mean latitude” steadily drifting equatorward, is challenged. Indeed, spots cluster around at as many latitudes as knots; a knot may appear at either lower or higher latitudes than previous ones, in a seemingly random way; accordingly, the spot mean latitude abruptly drifts equatorward or even poleward at any knot activation, in spite of any smoothing procedure. Preliminary analyses suggest that the activity splits, in any hemisphere, into two or more distinct “activity waves”, drifting equatorward at a rate higher than the spot zone as a whole.