"Nylon bag" technique, followed by 48 hours of chloridric pepsin digestion, was tried for determination of the digestibility of proximate nutrients of Rhodes grass hay (Chloris gayana, Kunth, cv. callide) and coconut caque meal, and compared with results from "in vivo" method. In a randomized blocks design, 100% of Rhodes grass, and mixtures of 85% and 70% of the grass hay with 15%, and 30% of coconut cake, and 100% coconut cake, made up four treatments (A, B, C and D, respectively). The coefficients of digestibility for Dry Matter and for proximate nutrients in treatments A, B and C, determined through nylon bags in bovine rumen, were compared to the coefficients determined by total faces collection, in sheep. It was observed and concluded that: a) Position of samples along the supoort-bars inside rumen, as well as dfferent periods of animal usage, had no influence on the results; b) The calculated proportions of coconut in the mixture to obtain best digestibility were 17.48% for Dry Matter, 21.10% for Ether Extract, 29.01% for Crude Protein, and 16.49% for Non Nitrogenous Extrat. Crude Fiber digestibility decreased linearly as the coconut proportion in the mixture increased; c) "Nylon bag" and "in vivo" methods gave similar coefficients of digestibility for Crude Fiber for all treatments; for other nutrients, however, there were differences depending on treatments; d) TDN values for coconut cake meal, determined through "nylon bags" were 68.76% taking the average coefficients of digestibility from treatments B and C, and 68.08% from treatment D; "in vivo", the average from B and gave 71.06% of TDN; such a small difference in NDT suggests to have had no interference of foods on digestibility; e) NDT calculate for Rhodes grass hay was 58.38% "in nylon bags", and 44.81% "in vivo", when treatment A was the only one considered; thus, it could be concluded that nylon bag plus pepsin method did not work for Rhodes hay.
digestibility; nylon bag plus, pepsin technique; coconut caque meal