We have the following indirect implication of form equivalence classes:
| Implication | Reference |
|---|---|
| 407 \(\Rightarrow\) 14 |
Effective equivalents of the Rasiowa-Sikorski lemma, Bacsich, P. D. 1972b, J. London Math. Soc. Ser. 2. |
| 14 \(\Rightarrow\) 229 |
Variants of Rado's selection lemma and their applications, Rav, Y. 1977, Math. Nachr. |
Here are the links and statements of the form equivalence classes referenced above:
| Howard-Rubin Number | Statement |
|---|---|
| 407: | Let \(B\) be a Boolean algebra, \(b\) a non-zero element of \(B\) and \(\{A_i: i\in\omega\}\) a sequence of subsets of \(B\) such that for each \(i\in\omega\), \(A_i\) has a supremum \(a_i\). Then there exists an ultrafilter \(D\) in \(B\) such that \(b\in D\) and, for each \(i\in\omega\), if \(a_i\in D\), then \(D\cap\ A_i\neq\emptyset\). |
| 14: | BPI: Every Boolean algebra has a prime ideal. |
| 229: | If \((G,\circ,\le)\) is a partially ordered group, then \(\le\) can be extended to a linear order on \(G\) if and only if for every finite set \(\{a_{1},\ldots, a_{n}\}\subseteq G\), with \(a_{i}\neq\) the identity for \(i = 1\) to \(n\), the signs \(\epsilon_{1}, \ldots,\epsilon_{n}\) (\(\epsilon_{i} = \pm 1\)) can be chosen so that \(P\cap S(a^{\epsilon_{1}}_{1},\ldots,a^{\epsilon_{n}}_{n})=\emptyset\) (where \(S(b_{1},\ldots,b_{n})\) is the normal sub-semi-group of \(G\) generated by \(b_{1},\ldots, b_{n}\) and \(P = \{g\in G: e\le g\}\) where \(e\) is the identity of \(G\).) |
Comment: