41 SUMMARY This study has been targetted to solve the difficulties to which the NPRP ( National Potato Research and Training Project ) face during the evaluation of segregating material obtained mainly from CIP ( Centre of International Potato Lima - PERU ). Material introduced from CIP as True Patato Seed were evaluated in Bozdağ - İZMİR and Menemen - İZMİR where the NPRP breeding studies run intensively. Different plant and tuber characteristics were evaluated on seedling progenies and on progenies of tubers which were obtained from same seedling progenies. Seed tuber sets which were used as seed tubers for obtaining tuber progenies, were prepared as genetically identical. Relation of the selected characteristics which are used as selection criteria in NPRP programs, were examined in and beetween the production levels and seasons. The relation of different characters in the seasons and of same character beetween the seasons were determined by using the data obtained from individuals. The data belongs to different characters were subjected to variance analysis seperately and combined on different progenies and the seasons to determine statistical parameters. According to the results of the analysis, although no significant difference was found beetween the family averages of seedling progenies and tuber progenies, the significant differences were found beetween productipn seasons of tuber progenies. Bozdağ as main crop production seasons was found significantly different and owercoming to the other seasons for the character of plant height, leaf area, number of tubers per plant, average tuber weight and tuber yield per plant. For all the characters, the different tuber progenies production seasons, except leaf area in Menemen Spring and Autumn seasons, were found significantly different from eachother.42 In the evaluation of the relation beetween characters, the relations were found relatively higher and significant in the calculation of data of genotypes than using family averages. This may happen, beside the differences of the distribution of data obtained from individual genotypes in each families, from one family to another. By depending on this opinion, it may be concluded that using measurements of individuals can give more accurate results than family averages, in this type of analisis. Since positive and significant relations beetween seedling progenies and tuber progenies were found for the skin color, this character can be considered that it may be used as selection criterium at seedling level. For the other characters especially the ones directly related with the yield, did not show significant relation beetween seedling and tuber progenies. So, it was found that using these characters as selection criteria at very early stage might produce a very high risk of the erosion of valuable materials. Beetween the seed tuber progenies production seasons, beside the skin color only number of tubers per plant had significant relations. Morower, this character showed positive and significant relations with yield and average tuber weight in all the seasons.

61