ConclusionsFrom the discussion of the results from 12 birds over three seasons it appears that much further work remains to be done. The results of this experiment are descriptive of what occurs with normal birds, but further experiments are clearly necessary to find out why. One of the major points is the variability between individual birds and this and not the statistical averages requires closer attention.For example, it has been known for a long time that shell thickness decreases in summer and it is almost certainly a temperature effect, but from the present results for individual birds, it is evident that whereas one bird exhibits this effect, another does not. What is the factor present in the second bird and absent or present in a smaller degree in the first ? If eggs with thinner shells tend to crack more easily, then birds laying thick-shelled eggs are a better commercial proposition from this aspect, but the ability of a bird to resist the tendency to produce thinner shelled eggs in summer may be equally important.Similarly, it has been shown that not all eggs of a given thickness will crack and more work is required on this problem.The next stage in relation to all characteristics considered in this paper seems to be experiments in which environmental factors such as temperature and humidity are controlled so that changes due to environment can be separated from genetic effects, and relationships between any two characteristics be more easily seen.Procedures for the production of a semi-synthetic, copper-deficient rat diet are described. The final product contains between 0.3 and 0.4 p g . of Cu per g. When supplemented with copper, the diet promotes rapid growth and good reproductive performance. With slight modification the diet would provide a suitable basal ration for studies of deficiencies of iron, zinc and manganese.