When, in nutritional studies, it is necessary to determine the losses of a particular nutrient in the faeces, two techniques are employed to ensure that the faeces collected are representative of the intake of food. The first consists of feeding a marker substance, such as carmine, immediately before and at the end of the experiment. This allows a separation of the faeces produced from the food given in the interval, and is the technique employed with non-ruminant species.With ruminants, however, the use of markers does not result in a clear separation of the faeces marking the beginning and end of the period. Consequently, the method employed to obtain accurate digestion coefficients with ruminants is to give the experimental ration in exact quantities for long periods, in order to ensure that a 'steady state' of faecal excretion is reached, and then to collect the faeces excreted during a measured interval of time. If refusal of food occurs on a particular day, the method is invalid, and so the amount of food given is usually kept small to avoid refusal. The low planes of nutrition which are thus employed are not desirable if results are to be applied in practical feeding (Schneider, 1954). Nor can the method be used where intake is voluntary or uncontrolled, because the faeces produced during a particular 24 h period are not simply referable to the food intake in some previous 24 h period.In the course of experiments designed to study the effect of methods of fodder preparation and nutritional plane on the energy metabolism of sheep, information was obtained on the passage of food through the digestive tract, its digestibility, and the diurnal variation in the amounts of dry matter and water excreted. This information is presented here, and the results have been examined in some detail in order to elucidate the problems involved in the determination of digestibility coefficients in ruminants. It has been found that the time relationships involved in the passage of food through the gut give rise to equations which allow the prediction of the errors of digestibility coefficients, the digestibility coefficients themselves and the diurnal pattern of faeces excretion. They also permit the estimation of the gut contents or 'fill' of the animal under a variety of feeding systems and show what precautions must be taken in the conduct of digestibility trials.
EXPERIMENTAL
Animals and technique for studying digestibilityThe six wether sheep used as experimental animals were harnessed for the separate collection of faeces and were housed in individual cages. Each sheep was given 600,