The rooted tree of life emphasizing key novelties and synapomorphies. Thumbnail sketches show major variants in cell morphology (microtubular skeleton red; peptidoglycan wall brown; outer membrane blue). The most likely root position is as shown; the possibility that it may lie within Chlorobacteria instead cannot yet be ruled out. Lowest level groups including or consisting entirely of photosynthetic organisms are in green or purple. The frequently misplaced hyperthermophilic eubacteria are in red; indel analysis confirms that Aquifex is a very divergent proteobacterium . The new negibacterial infrakingdom Gracilicutes segregates four phyla from the other negibacteria. Planctobacteria probably lost or reduced murein twice, as free-living Verrucomicrobia have murein. Note that 12 synapomorphies support the earliest branching of Chlorobacteria. The fact that mitochondria were present in the cenancestral eukaryote and that their ancestors, α-proteobacteria are a relatively recently derived of the eubacterial phylum Proteobacteria, proves that eubacteria must be significantly older than eukaryotes and decisively refutes suggestions that eubacteria may be derived from eukaryotes. As α-proteobacteria are nowhere near the root of the tree (irrespective of whether it is rooted beside or within chlorobacteria or as some mistakenly think between neomura and eubacteria) eukaryotes are substantially younger. The age of ~900 My for eukaryotes is based on a recent Bayesian analysis of 143 proteins multiply calibrated from the fossil record  and my own critical interpretation of the direct fossil record . This tree, though constructed from rare discrete cladistic characters, is remarkably similar to a 31-protein, 191 species universal sequence tree published while this paper was being reviewed ; see responses to comments by referee 3 for discussion of the few differences, all but one (the position of Aquifex) in regions poorly supported on the sequence tree.