Resumo:
An experiment was started in 1931 for the purpose of making a comparative study of the productivity of following varieties of Coffea arabica L, : Nacional (Coffea arabica L. var. typica Cramer) Amarelo de Botucatu (Coffea arabica L. var. typica Cramer forma xanthocarpa (Caminhoá) Krug) Bourbon (Coffea arabica L. var. bourbon (B. Rodr.) Choussy) Bourbon amarelo (Coffea arabica L. var. bourbon (B. Rodr.) Choussy forma xanthocarpa Krug) Sumatra (Coffea arabica L. var. typica Cramer) Maragogipe (Coffea arabica L. var. maragogipe Hort. ex Froehner) The results obtained from this experiment for the period 1935-1938 have been the subject of a previous publication. The present paper concerns results obtained from the same experiment for the period 1939-1946. The data obtained have been analysed by varieties and years in relation to : date of ripening of coffee berries, average size of coffee beans and total production. The results of this analysis showed that there were definite annual variations in production by varieties. With respect to ripening of fruits (cherries) on the varieties tested, it was found that the Amarelo de Botucatu variety had the earliest date of ripening in every year. The Nacional and Sumatra varieties varied by years but they were always later than the Amarelo de Botucatu in date of maturity of the cherries. The fruits of Bourbon and Boubon Amarelo were slightly later than all of of the above mentioned varieties and the Maragogipe matured fruit later than all other varieties studied. The data on the average size of coffee beans showed that the Maragogipe variety produced the largest bean. The Bourbon Amarelo and Bourbon produced beans of the smallest size. The Nacional, Sumatra and Amarelo de Botucatu produced beans that were intermediate in size between the Maragogipe and Bourbon. The analysis of production showed that for each year for the first four years (1935-1938) there was a definite yearly increase in production for each variety tested. The data obtained for subsequent years (1939-1946) show that, for all except the Maragogipe variety, each year with a high yield was followed by a year with lower coffee production (see figure 1). On the basis of the total production for the entire period studied (1935-1946) it was found that Bourbon Amarelo variety was the most productive followed closely by the variety Bourbon. The Sumatra variety was intermediate in total production and appreciably less than the Bourbon types. The lowest yielding varieties in this experiment were Nacional, Amarelo de Botucatu and Maragogipe. The production data have also been studied by four year periods and as a result certain interesting minor variations in relation to varieties have been noted. A more detailed statistical analysis of the production data from this experiment has been carried out by Prof. W. L. Stevens (see following article in this issue). He has showed that the coffee varieties studied can be placed in three statistically different classes. The varieties, by classes, according to their decreasing capacity of production, are as follows : a ) Bourbon Amarelo and Bourbon ; b) Sumatra ; c) Nacional, Amarelo de Botucatu and Maragogipe.
