American Scientist

To the Editors:

In "Dating Ancient Mortar," by John Hale, Jan Heinemeier, Lynn Lancaster, Alf Lindroos and Åsa Ringbom (March– April), the premise that carbon-14 is fixed in limestone-derived substances at the "exact time of construction," appears to seriously underestimate long-term chemical alterations of ancient mortars and plasters. Scanning electron microscopy of comparable samples from the pre–79 a.d. Roman town of Pompeii indicates that lime plaster hardening is not restricted to the few hours or days after the plaster dried. Instead, the development of extensive and interlocking lath-like calcium carbonate crystals in Pompeian plaster (Figure 1) is the result of an ongoing process. The authors' use of radiocarbon analysis of ancient plasters and mortars could not differentiate between ¹⁴C in the initial lime mix and ¹⁴C introduced later by atmospheric CO2 associated with calcium carbonate recrystallization.

Peter Grave
University of New England
Armidale, NSW Australia

Dr. Hale and his co-authors reply:

All mortar samples are not alike. Some present problems of long-lasting chemical activity with abundant recrystallizations. As we indicated in our article, for a number of Roman samples "the correct date was indicated by the second rather than the first fraction of carbon dioxide released in the analysis, because the mortar dissolves slowly but contains rapidly dissolving contaminants." The young calcite crystals shown in the scanning electron micrograph sent in by Dr. Grave fall in this category of contaminants.

In such cases, we plotted the percentage of CO₂ extracted from the sample as successive fractions over a period of about 16 hours. The resulting curve or profile raised the question: Which fraction or fractions match the correct historical age? In many mortar samples containing beach sand, the first fraction provided the match. However, with hydraulic and pozzolana mortars as well as contaminated samples, the ¹⁴C values stabilized around the historically correct date in the middle fractions, after 20 to 40 percent of the sample had been dissolved. Young calcites and recrystallizations dominate the first fractions, yielding dates that are too recent, whereas dead carbon from unburnt limestone residues dominates the last fraction, yielding dates that are too old. We also plotted the values of O₂ and ¹³C as the samples dissolved. The values themselves did not accurately reflect the age of the sample, but the point at which these two profiles leveled out showed that the equilibrium stage had been reached, thus helping to identify the fractions with correct ¹⁴C values.

We believe that if Dr. Grave applies this method to his mortar samples from Pompeii, he will obtain ¹⁴C dates that match the known historical ages of the Roman buildings.