Previous attempts to produce nonenzymatic template-directed oligomerizations of activated pyrimidines on polypurine templates have been unsuccessful. The only efficient reactions are those where the template is composed primarily of pyrimidines, especially cytosine. Because molecular evolution requires that a synthesized daughter polynucleotide be capable of acting as a template for the synthesis of the original polynucleotide, the one-way replication achieved thus far is inadequate to initiate an evolving system. Several uracil analogs were used in this investigation in order to search for possible replacements for uracil. The monomers used in this investigation were the imidazolides of UMP, xanthosine 5'-monophosphate, the bis-monophosphates of the acyclic nucleosides of uracil, and 2,4-quinazolinedione. The concentrations of various salts, buffers, pH, and temperature were among the different variables investigated in attempts to find conditions that would permit template-directed oligomerizations. Although the different monomers in this study demonstrated varying abilities to form very short oligomers, we were unable to detect any enhancement of this oligomerization that could be attributed to the poly(A) template. Although special conditions might be found that would allow purine-rich templates to work, these reactions cannot be considered robust. The results of our experiments suggest that pyrimidines were not part of the original replicating system on the primitive Earth. It has already been shown that ribose is an unlikely component of the first replicating systems, and we now suggest that phosphate was absent as well. This is due to the low solubility of phosphate in the present ocean (3 x 10(-6) M), as well as the difficulty of prebiotic activation of phosphates.