As progênies de cafeeiros murta são constituídas de plantas bourbon (ttNaNa), murta (ttNana) e anãs (ttnana), na proporção de 1:2:1. A fim de comparar as produções de cafeeiros bourbon e murta, progênies de 14 plantas matrizes murta foram estudadas em Campinas, cada progênie sendo inicialmente constituída de 10 plantas murta e 10 plantas bourbon. Em Pindorama estudaram-se apenas cinco progênies, constituída cada uma delas de cinco plantas murta e cinco bourbon. Tôdas foram plantadas a um pé por cova, para estudo das produções individuais. As plantas anãs não foram levadas para o campo por serem de desenvolvimento excessivamente lento e apresentarem produção pràticamente nula. Algumas delas foram conservadas em viveiro a fim de se observar seu desenvolvimento e frutificação. Verificou-se maior número de falhas entre as plantas murta do que entre as bourbon, principalmente em Campinas, indicando que os cafeeiros murta são mais fracos do que os cafeeiros bourbon. Em Pindorama esta diferença foi menos acentuada. A produção total média no período 1939-1952 em Campinas e em Pindorama foi analisada, confirmando-se os resultados já anteriormente obtidos, em Campinas, de que a produção do conjunto de plantas bourbon é bem maior do que a das plantas murta. A parte bourbon produziu cêrca de 70% a mais em Campinas e 100% a mais em Pindorama, do que a parte murta. O alelo na, além de influir sobre alguns caracteres morfológicos, deve ser o principal responsável pela drástica redução da produtividade das plantas murta. Os efeitos do alelo na não se fazem sentir no tamanho e peso dos frutos e sementes. É mais provável que a ação dêste alelo se relacione com a redução do número total de flores nas plantas murta ou no pegamento destas flôres, resultando em um menor número de frutos, em relação às plantas bourbon.
The murta variety of Coffea arabica is a heterozygous ttNana, its progeny segregating bourbon (ttNaNa), murta (ttNana) and nana (ttnana) in the approximate ratio of 1:2:1. In order to observe the effects of the na allele on yield several selfed progenies from murta individuals were studied simultaneously at the experiment stations of Campinas and Pindorama. In the first of these locations 14 progeny rows of 20 plants each were planted, each progeny having initially 10 bourbon and 10 murta plants. At Pindorama only five progeny rows were planted, each of which had an initial number of five plants of each type, giving a total of 50 individuals. Dwarf (nana) plants were not transplanted to the field due to their growth habit and to their very low yield. A few dwarf plants, however, were kept in a lath-house so that their development could be followed. A few years after planting the progenies in the field it was noticed that a proportionally higher number of murta plants died from the eifects of die-back. The final percentages of misses among the murta individuals were 34 per cent at Campinas and 21 per cent at Pindorama. Missing plants among the bourbon individuals of the progenies at Campinas were much less frequent, that is, only three per cent, contrasting with 19 per cent at Pindorama. It was also observed that death of murta plants occurred more frequently in certain progenies in both locations. The average yield for the period 1939/52 was studied for the progenies at Campinas and at Pindorama. It was observed that the average yield of the bourbon plants in every progeny was much higher than the yield of the corresponding murta group. The average yield of all bourbon groups exceeded murta by 70 per cent at Campinas and by 100 per cent at Pindorama. These results are in agreement with the conclusions of a previous study based on the first eight consecutive yields. It seems that such a large difference in yielding ability of these two genotypes can be ascribed mainly to the effects of the na allele, which also has a marked pleiotropic effect on vegetative characteristics, particularly when in a homozygous condition. In the heterozygous condition the na allele does net affect the size of flowers, fruits and seeds, while in homozygous individuals the flowers are slightly smaller, although the size of fruits and seeds are similar to those of bourbon. Several instances of somatic mutations have been found in ttnana plants giving origin to phenotypically murta (ttlNana) branches. When these mutated branches are allowed to remain they tend to overgrow the nana parts and to yield regularly, while the nana parts rarely produce flowers and fruits. This seems to corroborate the conclusion that the na allele has a major effect in reducing the yield of murta plants. Since this gene has no visible effect on fruit and seed size, it may be assumed that the lower yield of murta plants is due to a reduction in flower number and/or to a smaller fruit setting.