On the geometry of unitary involutions (Q747967)

Using a geometric approach the authors study the involutions of the group \(P\Gamma U(n+1,q^2)\). Specifically they prove the following: Let \(\theta\) be an involution of \(\mathrm{PG}(n,q^2)\) fixing a non-degenerate Hermitian variety \({\mathcal H}(n,q^2)\). Let \({\mathcal U}\) denote the unitary polarity associated with \({\mathcal H}(n,q^2)\). If \(\theta\) is linear and \(q\) is odd then either \(\theta\) fixes pointwise a non-degenerate \(k\)-dimensional section of \({\mathcal H}(n,q^2)\), \((n-1)/2 \leq k \leq (n-1)\) and its conjugate subspace with respect to \({\mathcal U}\) or \(n\) is odd and \(\theta\) fixes pointwise two skew generators of \({\mathcal H}(n,q^2)\). If \(\theta\) is linear and \(q\) is even then \(\theta\) fixes pointwise a \(k\)-dimensional subspace contained in \({\mathcal H}(n,q^2)\), \(0 \leq k \leq (n-1)/2\) and its conjugate subspace with respect to \({\mathcal U}\). If \(\theta\) is non-linear and \(q\) is odd then either \(n\) is odd and \(\theta\) fixes pointwise a symplectic Baer subgeometry of \({\mathcal H}(n,q^2)\) or \(\theta\) fixes pointwise a non-degenerate Baer quadric of \({\mathcal H}(n,q^2)\). If \(\theta\) is non-linear and \(q\) is even then either \(n\) is odd and \(\theta\) fixes pointwise a symplectic Baer subgeometry of \({\mathcal H}(n,q^2)\) or \(\theta\) fixes pointwise an \((n-1)\) dimensional Baer subgeometry of \({\mathcal H}(n,q^2)\) contained in a degenerate or non-degenerate hyperplane section of \({\mathcal H}(n,q^2)\) depending on the parity of \(n\).