Indian Council of Agricultural Research
*The paper summarizes the results of experimental investigations on cotton (G. arboreum) initiated by me at the Institute of Plant Industry, Indore, in 1940, in order to elucidate and verify certain aspects of the theory of polygenic inheritance. Some of the results have been discussed in the annual reports of the Institute and occasional papers. Part or the investigation, especially the two-way selection experiment, is being continued. For various reasons, the present time was considered opportune for bringing together in a connected account the principal findings of the analysis made so far. It is hoped that the results will be of interest to plant breeders as well as geneticists, as the experimental material, namely intraspecies crosses in cotton, was within the range of breeding material commonly employed for crop improvement. —V. G. Panse.
The paper deals with an experimental study of quantitative inheritance. The data were collected at the Institute of Plant Industry, Indore. Crosses had been made between eleven superior and nine inferior Bengals strains of cotton, G. arboreum. Progenies from these were grown in replicated plots in the subsequent generations. Selection had been practised for high and low fibre length from F3 onward in six crosses.
The material shows definite heterosis, as measured by the increase of F1 values over the mean of the parents. The relative efficiency of the different strains for crossing has also been studied. The genetic variability present in the material has been estimated and it has been shown that for crosses with a relatively low coefficient of variation, as in the present case, attention should be concentrated on mean values for making further selection. Crosses for selection may be chosen on the basis of their F1 values. The components of phenotypic variation have been estimated and it has been found that the estimated values are negative in some crosses. The only explanation that can be offered is that these estimates are subject to large sampling errors. The number of factors have been estimated by Panse's and Student's methods. It has been shown that for materials where selection is practised in two opposite directions, high and low, the estimates of the number of factors can be obtained relatively easily by the latter method. The response to selection has been studied and it is seen that although there is a continued response in the high lines, the response has come to a standstill in the first two or three selected generations in the low lines. A tendency for increase in the values of low lines with continued selection was ascribed to the positive correlation observed between yield and fibre length.
