Abstract
Mather''s theory of correlated responses postulates that selection for a normally balanced, polygenically controlled character produces unusual re-combinants from crossing-over. Other changes occur, and variations in other characters thus released, owing to mingling of different polygenes along the chromosomes. Animal and plant breeding selection experiments showing correlated responses are: analysed in the light of this theory. Selection for (1) the hooded pattern in rats led to sterility; (2) high and low body weights in mice produced psychological,morphological and physiological differences; (3) resistance to tumor formation in mice, following methylcholanthrene injections, led to a kaleidoscope shift in types of induced tumors; (4) high and low abdominal bristle number in Drosophila produced changes in other bristles, in spermathecae pigmentation, and in mating behaviour. There was gradual reduction in fertility as selection proceeded. Selection for: (1) high and low oil content in maize produced morphological and agronomic differences, and became hindered by decreased fertility; (2) earliness and cold tolerance in sweet corn led to pseudo-starchiness; (3) pleiocotyly in outbreeding species produced correlated changes in vigor, morphology and disturbances in hormonal balance, and increased lethality. Inbreeding species so far have shown no correlated responses. The transformation of pneumococcal types and changes in viruses during their passage through plants are also related to the theory. Some evolutionary aspects of correlated heredity are discussed. Extinction of some animal and plant species may have occurred by specialized selection bringing about correlated loss of fertility. An evaluation is outlined of correlated responses in plant breeding, especially for pleiocotyly selection. Their disadvantages will not be felt providing lop-sided selection is not practiced to an extreme.