Sampling Fecal samples were collected every 4 hours for the first 40 hours after dosing in order to determine the time at which particles first appeared in feces. After a small aliquot was removed for analysis, the remainder of the fecal material was saved for compositing into a 24-hour sample. The 4-hour samples were mixed for 10 minutes with a Hobart mixer before aliquots were taken. The 24-hour samples, made up of 4-hour collections during the first 40 hours and fecal collections made at milking time thereafter, were also mixed with a Three aliquots of each 24-hour fecal sample were taken for Hobart mixer. analyses. The cows were transferred to individual outdoor pens with concrete floors 7 Since some activity was still detectable in feces, grab days after dosing. samples of fresh fecal material were collected for 9 additional days. During this period, the daily particle output was estimated from the grab sample concentrations and the average fecal output rates for each cow. Samples of urine, milk, and blood were collected 24 hours after dosing in order to check leaching from the particles and subsequent gastrointestinal uptake of the radionuclides. Analyses All samples were analyzed by gamma-ray spectrometry using a Ge(Li) detector (germanium-lithium drifted) and a pulse-height analyzer. Samples were counted so that the 20 counting error was less than 10 percent or the counting time exceeded 80 minutes. The minimum detectable activity in the samples with <10% error at the 95% confidence level following 80-minute counting times was 2.2 nCi for *®sc, the isotope with the lowest counting efficiency. This amount is less than 3 x 1073% of the administered dose and represents less than 0.4% of the administered dose in a typical 30-kg 24-hour fecal collection. About 10% of the low activity samples were counted for 1,000 minutes. These samples included feces taken when isotope levels had decreased to very low levels, plus the blood, milk, and urine samples taken at 24 hours after dosing. The minimum dectable *®Sc activity following 1,000-minute counts was 0.68 nCi at the 95% confidence level with 10% error. This amount is less than 10-3% of the administered dose and represents less than 0.1% of the administered dose in a typical 24-hour fecal collection. Data given in the results section for 24-hour fecal samples represent the radioactive-decay-corrected average for the three aliquots from each collection. RESULTS AND DISCUSSION Initial Excretion of Particles Particles, of each size were detected in feces from all cows at 12 to 16 hours after dosing. For the 20-ym particles, the highest concentrations in feces 155