9508987 Hillis Biologists have long been interested in the history of life, and the last few decades have been accompanied by major expansion in the areas of biology concerned with reconstructing these histories. There are now many different methods that can be used to reconstruct history, and there have been several new applications of the methods, such as discovering the US hantavirus and its animal reservoirs, as well as helping to determine whether health professionals infected their patients with HIV. An inherent difficulty in this important field of science is that there is usually no way to truly know the histories of the organisms being studied, hence it is difficult to know hmw often reconstruction methods fail. In the study proposed here, we will overcome that limitation. A virus will be grown in the laboratory to generate lineages with known ancestries. DNA sequences and other characteristics will be obtained from the resulting lineages. The success of reconstruction methods will be tested directly by comparing the estimated histories with the known histories. The data will also be studied for unusual properties that may thwart recovering history, such as the possibility of molecular convergence resulting from common selective pressures. %%% This study should impact the field of systematic biology both in establishing an empirical methodology for experimental phylogeny construction and in providing direct tests of phylogenetic inference methodology. ***