It has been suggested (Temple, 1977) that an obligatory mutualism existed between the extinct dodo (Raphus cucullatus) and the endangered tree [Calvaria major according to Owadally (1979), but actually a composite of two species (Sideroxylon sessiliflorum and S. grandiflorum) according to Friedmann (1981)]. Although some of the evidence on which that hypothesis was based is disputed (Owadally, 1979; Temple, 1979), the apparent near absence of natural germination and the thickened structure of the endocarp of the suggest that the tambalacoques (sensu lato) are very specialized plants, requiring some mechanism of endocarp abrasion to facilitate germination. Temple (1977) suggested that the dodo's gizzard accomplished this, and that since that bird's final extirpation in 1681, there has been no natural germination. dne test of Temple's hypothesis would be an accurate aging of the remaining trees, especially since there is disagreement between Temple and Owadally as to the age of those trees. They suggest the are 300 and 75-100 years old, respectively, even though their information apparently came from the same source. Friedmann (1981) has recently aged one specimen at between 30 and 50 years, and if other are significantly less than 300 years of age, any dodotambalacoque mutualism would at best have been facultative. However, the aspect of the Temple-Owadaily debate that is most surprising to me is the total neglect of an even more logical group of organisms in response to which the thickened tambalacoque endocarp may have evolved. These are the testudinid tortoises of the genus Cylindraspis (Bour, 1984, 1985; Geochelone, according to some authors). The significant enhancement of germination of the Galapagos tomato (Lycopersicon esculentum; Rick and Bowman, 1961) and the prickly pear (Opuntia sp.) by passage through the digestive tracts of the Galapagos tortoise (Geochelone elephantopus), and Berlandier's tortoise (Gopherus berlandieri; Rose and Judd, 1982), respectively, have already been documented. In addition, of many Aldabra Island plant species germinate readily after passing through the gut of the Aldabra tortoise [Geochelone gigantea (=Dipsochelys elephantina according to Bour, 1984); Hnatiuk, 1978] and germination may even be enhanced in some (Stoddart and Savy, 1983). Further, tortoises were undoubtedly abundant herbivorous inhabitants of Mauritius through the Pleistocene (Bour, 1979, 1985; Pritchard, 1979) and until their extirpation in the early 19th century (Auffenberg, 1974; Pritchard, 1979). Closely related species on Rodrigues Island (now extinct) were even observed eating apples, dates, and seeds from trees 300 years ago (review in Bour, 1981). There is, however, considerable difference of opinion as to how many tortoise species lived on Mauritius into Recent times. At least one species apparently went extinct on Mauritius shortly before humans arrived, and another shortly after (Auffenberg, 1974; Pritchard, 1979). In all, at least seven PleistoceneRecent tortoise species have been described from Mauritius (Auffenberg, 1974), but it is currently believed that these were variants (primarily sexual and ontogenetic) of only two (or possibly three) species (Auffenberg, 1974; Bour, 1979, 1984). Ecologically it is very interesting that two (or more) herbivorous tortoise species could coexist on Mauritius, an island of only 1865 km2. They must have had very different habits as Arnold (1979) suggested was the case for the two closely related tortoise species living on nearby Reunion Island. Based on morphology he speculated that one species (more saddlebacked) occupied more open habitats and browsed higher than the other (more dome-shelled) species. Similar differences probably also characterized the species on Mauritius. It is therefore just as logical to suggest a tortoise-tambalacoque seed coevolution as one involving the dodo. Experimentation with tambalacoque seed germination following mastication and passage through tortoise guts (for example, those from nearby Madagascar or the Aldabra islands in the Seychelles Archipelago) is clearly warranted. Enhancement of germination in such experiments would argue strongly in favor of the natural existence of a tortoise-tambalacoque mutualism prior to the former's disappearance. It would also provide the Mauritius Forest Service with a more natural alternative to the mechanical seed abrasion now used to induce