A collaborative study on the determination of available lysine in feeding stuffs

The methods of Carpenter 1 and Roach, Sanderson and Williams 2 for the determination of available lysine have been compared in two collaborative studies involving eight laboratories. The feeding stuffs used were: fish meal and groundnut meal in Trial 1 and soya bean meal, sunflower meal, a commercial mixed feed and wheat meal in Trial 2. There was no statistically significant difference between the two methods in terms of their overall means for any of the materials studied, though for mixed feed the values by Carpenter's method tended to be smaller. Inter‐laboratory standard deviations for available lysine were generally greater by Carpenter's method, but significance was established only in the case of wheat. ‐ Some of the problems encountered in both methods are discussed.