Time management is a key feature needed in any query language for web and semistructured data. However, only recently this has been addressed by the Semantic Web community, through the study of temporal extensions to RDF (Resource Description Framework). In this paper we show that the ability of the RDF data model of handling unknown resources by means of blank nodes, naturally yields a rich framework for temporal reasoning in RDF. That is, even without knowing the interval of validity of some statements we can still entail useful knowledge from temporal RDF databases. To take advantage of this, we incorporate a class of temporal constraints over anonymous timestamps based on Allen's interval algebra. We show that testing entailment in temporal graphs with the constraints proposed reduces to closure computation and mapping discovery, that is, an extended form of the standard approach for testing entailment in non-temporal RDF graphs.