form of the decay law, the time interval between ‘the explosion end collection dates, and the time interval between collection and counting. . Radioactivities on the gummed papersgiven in this report are corrected to the date of sampling, and recorded in the units "disintegration per minute per square foot per day", referred to ‘hereafter es td/m", sin filter ectivity, also corrected to the date of. sampling, is expressed in "disintegration per minute per cubic meter of filtered air" and is abbreviated Nd/n/n3". 1b SOTRCES - or ERROR in RADIOLOGICAL DATA In order to interpret ‘the radiological measurements, sources . of error must.be appraised. Any series of measurements of this ~ nature must include inaccuracies, some inhsrent in the methods, others the result of human error. ‘The sources of error and limitations of the data can be considered in two categories: “errors in the radiological data, and the representativeness of the gummed paper Samples as a of the ground contamination. 1.4.1 . measure . Observers! Errors The most frequent error appears to be misdating the | data cards. It is also possible that on some occasions gummed ‘papers were mailed with the wrong card. Errors originating at the collection station cannot, in general, be corrected. 1.4.2 Tabulation Errors. . Mistakes are known to occur in punching the ITEM ‘cards which were used for computing and tabulating the data. A recheck of questionable observations resulted in the correction of a large nimber of such errors. Tabulation error undoubtedly remain, but they comprise a very small fraction of the data. 1.4.3 Contamination of Samples in Processing In processing such a large number of samples by routine procedures, cross-contauination is possible, especially if a very active sample is present. Evidence that such errors exist can be seen from an examination of samples processed on Lo . gd fea - s - ‘ 1: November 1952, the date on which a very active gummed paper