Methods for the physical organization of temporal databases are discussed in this paper. Most of the known storage structures for temporal databases are based on one-dimensional access methods. An exception is the TSB-Tree structure [1], that takes into account special features of temporal dimensions of data and provides the execution of various types of queries. A variant of TSB-trees based on trie-hash [2] is described. The data storage structure suggested outperforms TSB-trees in terms of speed when performing searches in the current part of the database, while it ranks below TSB-trees in access speed to the archival part.