Abstract

                                Very often plant breeders face the difficulty of placing various types or cultures in order of merit for one or more of their attributes. Cotton varieties, for example might be ranked for their vigour, squaring or disease resistance into definite groups as better than, equal to or poorer than the standard (the local strain cultivated). New varieties under trial for the first time might be ranked in order of merit without assigning any numerical measure of intrinsic worth to each type. Ranking arises naturally in cases where for lack of time, money or instruments, measurement of a characteristic is considered impossible. It is not possible for a cotton breeder, for example, to count the number of bolls or squares of every type particularly when he deals with large collection of material. He can only rank these types into certain definite groups. We are sometimes forced to have recourse to ranking methods even when measurements have been made, in order to reduce the labour of computation or to get an idea of the general trend.

Key words :