The estimation of linkage from the offspring of selfed heterozygotes

Five methods of solution are given of the statistical problem presented by typical linkage data. The example chosen shows the various errors into which the use of inefficient statistics leads. Of the efficient methods the method of maximum likelihood possesses the advantage that it may be applied directly to any analogous problem, and is related in a previously unsuspected way to the measure of discrepancyX 2. The product ratio method, for using which a table is provided, enjoys the practical advantages of other efficient solutions, and is in addition unaffected by differential viability, if this is caused by one factor only. The method of minimumX 2, unlike the other two, is laborious in computation, and seems to possess no special theoretical interest.