The effect of year, which includes effects of date and density, was based on the value for Marley Wood + 0.2 in 1960, and on 1/ Marley + 2/3 Great Wood for 1961-69. The effects are given in table A3. Habitat. Differences in habitat type, classified as "good" and "bad," within each area produced a small, but only just significant, difference (mean weighted difference 0.17 -0.08 eggs), which was not considered worthwhile applying as a correction factor in the present analysis. The classification into "good" and "bad" was made subjectively by the authors, and is not comparable to that of Krebs (1970), who used consistently occurring extreme values for clutch size as criteria for habitat quality and ignored middle values. Corrections applied. To allow for the effects of "area" and "year" the effects calculated above in tables A2 and A3 were substracted from each observed clutch size. To allow for the effect of the age of the female, 0.3 eggs was added to the clutch of a yearling and subtracted from that of an older bird.