Let $M\left(\alpha \right)$ be the Mahler measure of an algebraic number $\alpha$, and $V$ be an open subset of $ℂ$. Then its Lehmer constant$L\left(V\right)$ is inf $M{\left(\alpha \right)}^{1/deg\left(\alpha \right)}$, the infimum being over all non-zero non-cyclotomic $\alpha$ lying with its conjugates outside $V$. We evaluate $L\left(V\right)$ when $V$ is any annulus centered at $0$. We do the same for a variant of $L\left(V\right)$, which we call the transfinite Lehmer constant $L_{\infty }\left(V\right)$.Also, we prove the converse to Langevin’s Theorem, which states that $L\left(V\right)\>1$ if $V$ contains a point of modulus $1$. We prove the corresponding result for $L_{\infty }\left(V\right)$.

Let α be a totally positive algebraic integer of degree d, i.e., all of its conjugates $\alpha ₁=\alpha ,...,{\alpha }_d$ are positive real numbers. We study the set ₂ of the quantities ${\left({\prod }_{i=1}^d{(1+\alpha {²}_i)}^{1/2}\right)}^{1/d}$. We first show that √2 is the smallest point of ₂. Then, we prove that there exists a number l such that ₂ is dense in (l,∞). Finally, using the method of auxiliary functions, we find the six smallest points of ₂ in (√2,l). The polynomials involved in the auxiliary function are found by a recursive algorithm.

Suppose $A\in G{l}_d\left(\mathbb{Z}\right)$ has a 2-dimensional expanding subspace $E^u$, satisfies a regularity condition, called “good star”, and has $A^{*}\ge 0$, where $A^{*}$ is an oriented compound of $A$. A morphism $\theta$ of the free group on $\{1,2,\cdots ,d\}$ is called a non-abelianization of $A$ if it has structure matrix $A$. We show that there is a tiling substitution$\Theta$ whose “boundary substitution” $\theta =\partial \Theta$ is a non-abelianization of $A$. Such a tiling substitution $\Theta$ leads to a self-affine tiling of $E^u\sim {ℝ}^2$ with $A_u{:=A|}_{E_u}\in G{L}_2\left(ℝ\right)$ as its expansion. In the last section we find conditions on $A$ so...

The intrinsic algebraic entropy ent(ɸ) of an endomorphism ɸ of an Abelian group G can be computed using fully inert subgroups of ɸ-invariant sections of G, instead of the whole family of ɸ-inert subgroups. For a class of groups containing the groups of finite rank, aswell as those groupswhich are trajectories of finitely generated subgroups, it is proved that only fully inert subgroups of the group itself are needed to comput ent(ɸ). Examples show how the situation may be quite different outside...