recording of mortality in the two communities (without introducing new methods of data collection at this stage). Here the point is to be emphasized to our collaborator, that ascertainment of mortality must be identical in the exposed and control communities. That is, ties in the affected community, deaths not otherwise recorded, community, or discounted. extra health care activi- that might lead to finding must be balanced in the control It is expected that such prelimi- nary or planning investigations as an examination of mortality will have a beneficial educational effect in the circumstances. No more need be said at this time about a study of health effects, except to mention later, actual that the sampling frame discussed above would provide the basis for the health Study also (of course, a similar frame would require to be developed for any control population) . Summary of Recommendations Controversy can be avoided by rewriting the proposal to separate: ‘planning (i} determination of retention levels; (1i) for health studies not presently accessible or advisable; (ii) a possible future study of health. The greater part of available funds and energy should be channeled to determination of accurate levels of internal contamination. Educational and preliminary planning activi-~ ties to develop sufficient epidemiologic methodology for continuing, scientifically valid study of heaith, are fully justified. a 3009432 however,