The relationships between the 2:1 layer silicates are discussed in terms of the charge balance between the ions occupying X, Y, and Z sites. This brings out the relationship of glauconite to the illites, and clarifies the reason for the ideal occupancy of the Y sites in montmorillonite. It not only shows hectorite, saponite, and sauconite to be tri-octahedral smectites, but also shows vermiculite to be a tri-octahedral analogue of the illites. Apparently anomalous differences in permitted compositions of the di-octahedral and tri-octahedral micas are explained in terms of mismatch between the tetrahedral and octahedral layers, and it is shown that ephesite is to be regarded as a tri-octahedral common mica, not an analogue of margarite. Anandite is the nearest known naturally occurring equivalent to a tri-octahedral analogue of margarite, but the true analogue is a barium magnesium mica synthesized by Frondel and Ito.