Cohomological detection of Algebraic cycles

Abstract

The objective of this thesis is twofold. The first is to give simpler more unified treatment, for all the diverse examples of non-trivial nullhomologous cycles. Then this would allow to construct new examples of such cycles. The assumptions made have been checked in various papers in specific examples. The second is that it's been able to employ the method of detection explained in this work, to detect various of these null homologous cycles. More precisely it is checked with the exception of cycles over arithmetic varieties.