Proceeding with the publication begun in a previous paper, the authors report the results of further experiments designed to test the efficiency of various phosphates for cotton. This second series was conducted between 1941-42 and 1944-45. While one of the six experiments utilized to determine the effects on the yields was annual, the others were conducted in the same plots for two to four seasons. The yearly application of P2O3, 80 kilograms per hectare, was always used in the presence of nitrogen and potash. Except in one of the experiments, where no effect was observed, the responses to phosphorus varied from fair to very high. In the average of the latter five experiments, bone meal and Serranaphosphate gave similar results and were superior to Cibraphosphate, while superphosphate was superior to all of them. As a rule these differences tended to decrease as the number of years of the experiments was increased. In four of the five experiments phosphorus hastened the fruiting march of the cotton plant. In this regard the influence of the various phosphates was similar in two experiments, while in the others, located on soils highly deficient in this nutrient, superphosphate was superior. Generally phosphorus increased the weight of the bolls and seeds, but did not influence the lint percentage and length, even where the effects an the yields were very high. The residual effect of phosphorus, which was studied in two of the experiments, was very high, and even on the "terra-roxa-misturada" soil (one experiment) superphosphate, bone meal and Serranaphosphate gave similar results, these three being superior to Cibraphosphate.