ON TOOTH SUCCESSION IN DIADEMODON.

Abstract

Two characteristics clearly distinguish between the dentitions of typical mammals and reptiles: generally the former are heterodont and diphyodont while the latter are homodont and polyphyodont. Superficially it might seem that by progressively reducing the number of replacement teeth in an evolutionary sequence of animals, it was a simple matter to arrive at a diphyodont mammalian descendant from a polyphyodont reptilian ancestor. But data on the sequences in which teeth are initiated in mammalian and reptilian embryos suggest that the process was more complex. In most reptilian embryos teeth are probably developed in sequences which pass through alternate positions from the back to the front of the jaw (Osborn, 1971). It is not easy to relate this sequence to that observed in mammalian embryos; for instance for the upper jaw of Eremnitalpa, Osborn (1970) gives the sequence it, i2, C, m4/M1/m3/i3/M2/M3/I1, 12, C, P3, P4/ 13 (in which the lower case represents deciduous teeth and the sequence is known only where bars separate teeth). Both heterodonty and a very limited number of replacement teeth also evolved in the postcanine dentitions of the gomphodont cynodonts and their descendants, the tritylodonts of the Triassic and Jurassic (Crompton, 1973). Although these animals finally become extinct in the middle Jurassic, it seems possible that the manner in which they evolved their limited numbers of replacement teeth from the ancestral polyphyodont condition might shed light on the evolution of diphyodonty in mammals. The earliest of the well-known gomphodont cynodonts with a markedly heterodont postcanine denition and a linited number of replacement teeth was Diademodon. The purpose of the present paper is to plot the sequence in which postcanine teeth were replaced in Diademodon and to compare this with tooth replacement in a polyphyodont reptilian dentition. If both have been correctly understood it should indirectly be possible to obtain some idea of the biological controls which regulate tooth replacement. And it can be safely assumed that by varying similar biological controls, mammalian diphyodonty could also have been evolved from polyphyodont dentitions. The pattern by which Diademodon, a lower Triassic gomphodont cynodont, replaced its postcanine teeth has recently been studied by Crompton (1963), Ziegler (1969) and Hopson (1971), and is also referred to by Crompton (1973). In view of the fact that no new data is provided in the present paper, it might seem unnecessarily contentious to propose a fourth solution to this problem within eight years. However, during the last decade almost all workers have, with varying degrees of success (e.g., DeMar, 1972) attempted to fit patterns of tooth replacement into the framework of a general model known as the Zahnreike theory. This theory, which was proposed by Edmund (1960, 1962), apparently links together tooth replacement patterns encountered throughout the vertebrate kingdom. My own studies, referred to above and later, have convinced me that the Zaknreike theory as a model describing the initiation of patterns of tooth replacement is untenable in the face of the phylogenetic and embryological evidence. I have concluded that the rate of tooth replacement is partly controlled at each tooth position.