A model based on experiment and ab initio theory for the high-yield carbon--arc synthesis of C 60 and other fullerenes is presented. Evidence that is given indicates that the synthesis must start with the smallest units of carbon (atoms, dimers, etc.). The model is then broken into four steps: (1) the growth of carbon chains up to length C 10 from initial reactants present in the carbon vapor, (2) growth from chains into monocyclic rings (C 10 -C 20 ), (3) production and growth of three-dimensional reactive carbon networks (C 21 -C x , x = 30-40), and (4) growth of small fullerene cages via a closed--shell mechanism that exclusively produces C 60 , C 70 , and the higher fullerenes as the stable products.The carbon-arc synthesis of from solid graphite surely represents one of the most phenomenal phase transitions ever discovered. This synthetic technique, developed by Kratschmer, Fostiropoulos, and Huffman (7), has made bulk quantities of and other fullerenes available to the scientific community. This availability has stimulated a tremendous effort directed at characterizing this novel and exciting new class of molecules. One of the most fascinating aspects of the fullerenes, however, is the chemistry of their formation within the carbon arc. The synthesis is amazingly simple, and yet it produces results that are so fantastic and unexpected! The best experimental evidence (more on this later) indicates that the graphite reactant is vaporized in the carbon arc into the smallest units of carbon-atoms and possibly dimers-which then, through a concerted series of reactions, and in a limited range of pressures and temperatures, recombine to produce the spheroidal shells of carbon known as the fullerenes (2).How can this reaction scheme be understood? What are the individual chemical mechanisms that lead to the formation of the fullerenes? What reactive intermediates are produced in the carbon arc? How does the hightemperature chemical environment of the carbon arc produce the low-entropy