statistical significance to their findings. However, simply using the total number of samples in each of the river, raw, and finished categories, we have calculated the 95 and 99% confidence intervals for the probability of a transformation in each of the water types by applying binomial distribution theory. One would assume that the probability of a transformation in the control group is very small. If this probability is less than the lower limits of the intervals shown in Table 11, the particular water category can be said to have had cellular transformations at a statistically significant level (0.0275 and 0.005 levels for the 95 and 99% intervals, respectively) compared to the corresponding control group. Let us now compare results between groups. Two-way contingency tables comparing the observed results for the three classes of tested waters are shown in Table 111.The x2 statistic indicates that there is no significant difference between the number of cell transformations found using river, raw, or finished samples. Conclusion: neither conveyance of river water to treatment plants nor treatment itself reduces whatever unknown mechanisms cause the indicated transformations.
SIR: We appreciate the questions and statistical efforts of Kaplan and Thode with respect to our paper. An estimate of the expected number of transformations in the controls is given in the table taken from our paper by Kaplan and Thode. Fifty control cultures were serially propagated for 8 weeks with no cell transformation seen among the progeny. In addition, other studies with this cell line were carried out over a 2 yr period. If one assumes, as they have done, that a binomial distribution is appropriate, then the estimate of the variance and the confidence interval that does not overlap zero is significant; therefore, the intervals calculated by Kaplan and Thode in their Table I1 indicate that all types of water samples tested were significantly different from the controls at 95 and 99% confidence limits except finished water at the 99% level. Results between treatment groups as compared by Kaplan and Thode indicate no significant differences. The only proper comparison is between controls and any one of the three
The reported results would have been far more important if we had had additional information as to control samples. Furthermore, the authors allude to organic compounds as causing the transformations, substances we presume are present at very low concentrations whose detection requires sophisticated technology. ,It thus behooves the authors to provide further information as to the characteristics of the sampling sites (e.g., below sewage outfalls, downstream of industrial point sources of discharges, etc.), as well as the times of sampling (e.g., before or after chemical spills, hydrological conditions, etc.). Despite these shortcomings, the studies by Pelon and associates are important to our understanding of relationships between human health and water contaminants of a class significantly different from those we have faced in generations past. Edward Kaplan Henry C. Thode, Jr. Brookhaven National Laboratory Upton, New York 11973
treatments. The cell transformations may have been caused by the same or different chemicals. Our research did not attempt to determine precise causes of transformations since all three water types contain numerous organic and inorganic compounds. We made no statement that the cell transformations were caused by organic compounds, but we did state that organic compounds were present. Detailed information regarding point sources of industrial discharges of organic and inorganic .compounds within the sampling area may be obtained from the initial citation in our paper.
William Pelon David E. Lesley Department of Microbiology and Immunology Louisiana State University Medical Center 1100 Florida Avenue New Orleans, Louisiana 701 19
Correction
1980, Volume 14 Mendel Friedman,* M a r t i n J. Diamond, a n d J a m e s T. MacGregor: Mutagenicity of Textile Dyes. Page 1145. In Table I, compound 4 should be (2.1. Reactive 83.
122
Environmental Science & Technology
0013-936X/81/0915-0122$01.00/0 @ 1981 American Chemical Society