Three ultrafilters in a modular logic (Q429920)

The authors conjecture that the equational theory of the modular logics (i.e., modular ortholattices) is completely determined by that of the finite-dimensional modular logics. This strengthening of a conjecture of \textit{G. Bruns} [Houston J. Math. 9, 1--7 (1983; Zbl 0518.06007)] is closely related to attempts to obtain a full description of the coarse structure of the lattice of varieties of modular logics. As the authors note, if an arbitrary modular logic were embedded into an atomistic one, then their conjecture almost certainly would be solved positively. A natural method of embedding for complemented modular lattices is given by \textit{O. Frink} [Trans. Am. Math. Soc. 60, 452--467 (1946; Zbl 0060.05811)]. The main result of the paper under review is an example that shows that this approach fails: not all of the orthogonality structure of a modular logic can be carried through such an embedding. Namely, they present a modular logic \(\tilde L\) such that the Frink lattice \(\mathrm{Fr}(\tilde L)\) in which \(\tilde L\) is embedded by his method has an initial segment which is not orthomodular. (This is where the three ultrafilters mentioned in the title come in.)