curred subsequently to discharge to the lake, since mirexcontaminated sediments did not have a significant amount of 8-monohydromirex compared to mirex (2, 6 ) . Because a large proportion (>30%) of the mirex-related compounds in biota are photodegradation products, most of the mirex which entered the food web must have been held in the water column and recycled to the surface of the lake over a period of years. Alternatively, the rate of photolysis may have been enhanced by some unknown mechanism in association with dissolved organic matter, particle surfaces, or phytoplankton.The constancy of the ratio of 8-monohydromirex to mirex levels in Lake Ontario herring gul1,eggs over the years 1972 to 1978 is shown in Table IV. The ratio h samples of five species of fish (coho salmon, alewives, smelt, carp, and eel) from Lake Ontario in 1977 was also in the range 0.3-0.4 (6, 7). These data suggest that an equilibrium between input to the lake and photodegradation existed. However, the declines in mirex levels in herring gull eggs since 1974 ( 1 9 ) , presumably associated with cessation of direct discharge to the lake, should have resulted in an increase in the ratio if photodegradation had been occurring throughout this period. That portion of the mirex and photomirex in the lake which is available for bioaccumulation in fish and herring gulls must therefore be protected from further photodegradation by being sequestered and cycled between and within various food webs over a period of several years. The data also indicates that mirex in the sediments is not being recycled into the ecosystem at an appreciable rate.