The mechanism of cytoplasmic inheritance has been viewed in two ways which, though basically different, may not be mutually exclusive. From one standpoint, cell characteristics are determined by a self-perpetuating pattern of interrelationships among a variety of molecular species. Another more classical concept has been called by Nanney (1957) the Master Molecule concept. It assumes a special material acting as a reservoir of information which controls specific cell characters. Because some of this special material must be present in every cell, the importance of self duplicating particles, or units, is obvious. Chromosomes fit well into this scheme, as do certain cytoplasmic particles that can be traced from cell to cell. I will review briefly some of the evidence for the continuity of the centriole and say something about its possible genetic significance. Centrioles were first observed in animal cells nearly a century ago by van Beneden. He noted that these cytoplasmic particles divided before mitosis or meiosis and appeared to have the power to organize the mitotic apparatus and direct the migration of the separating chromosomes. Boveri (1887) then added greatly to interest in these particles by proposing the theory that the essential event of fertilization of the animal egg was the introduction into the egg of the centrioles of the spermatozoan. The centriole was thought to be duplicated at each somatic cell division and to be handed down continuously from generation to generation of cells. It was thus considered a permanent and autonomous cell organ that arose only by the division of a preexisting body of the same kind. Serious criticism arose against this concept regarding the importance of the centriole in cell division and the doctrine of the universality of its genetic continuity. For one thing, in many cells including the higher plants, no centriole could be seen at all, whereas in others a centriole was apparent during only a portion of the mitotic cycle. Furthermore, what appeared to be conclusive evidence of the de-novo origin of centrioles came from the demonstration by Morgan (1900), and later by others, that eggs could be experimentally induced to divide without the contribution of centrioles from the spermatozoan. Since it was considered that the egg's own centrioles were non functional, or were lost, at the time of egg maturation, the new centrioles that appeared in the so-called tcytasters of the par*Presented at a symposium, Cytoplasmic Units of Inheritance, sponsored by the American Society of Naturalists, and held in Montreal, Canada, December 27, 1964. The special engraving and printing of the half-tones were made possible by a grant from the National Science Foundation.