Some computations in equivariant cobordism in relation to Milnor manifolds

From MaRDI portal
Publication:2452035



Abstract: Let mathcalN∗ be the unoriented cobordism algebra, let and let Z∗(G) denote the equivariant cobordism algebra of G-manifolds with finite stationary point sets. Let epsilon∗:Z∗(G)omathcalN∗ be the homomorphism which forgets the G-action. We use Milnor manifolds (degree 1 hypersurfaces in RPmimesRPn) to construct non-trivial elements in Z∗(G). We prove that these elements give rise to indecomposable elements in Z∗(G) in degrees up to 2n−5. Moreover, in most cases these elements can be arranged to be in mathitKer(epsilon∗).


The Milnor manifolds \({\mathcal H}(m,n)\), \(m\leq n\), are defined by \[ {\mathcal{H}}(m,n)=\{([x_0:\cdots:x_m],[y_0:\cdots:y_n])\mid \sum_{j=0}^m x_jy_j=0 \}\subset {\mathbb R}P^m\times {\mathbb R}P^n. \] In this paper the authors consider the natural action of the \(2\)-torus \(G={\mathbb Z}_2^n\) on \({\mathcal H}(m,n)\) and show, using a criterion proven in [\textit{G. Mukherjee} and \textit{P. Sankaran}, Proc. Am. Math. Soc. 126, No. 2, 595--606 (1998; Zbl 0889.57041)], that it often defines indecomposable elements in the equivariant cobordism ring \(Z_*(G)\) of smooth closed manifolds with a \(G\)-action having finite fixed point set. On the other hand, forgetting the group action, it is known which Milnor manifolds are boundaries [\textit{S. S. Khare} and \textit{A. K. Das}, Indian J. Pure Appl. Math. 31, No. 11, 1503--1513 (2000; Zbl 0967.57029)], thus the result provides information on the kernel of the forgetful homomorphism \(Z_*(G)\to {\mathcal N}_*\) to the ordinary unoriented cobordism ring.











This page was built for publication: Some computations in equivariant cobordism in relation to Milnor manifolds

Report a bug (only for logged in users!)Click here to report a bug for this page (MaRDI item Q2452035)