Recently W. McCune and D. A. Sands have considered, in a note Amer Math. Monthly 103 (1996), 888{892], implicative axioms in groups of the form if u = v, then w = t, where u, v, w, t are group words in some variables. Their interest is in such identical implications that deene the variety of groups, or the variety of abelian groups. What I want to present is, slightly more generally, identical ...