1. In the first of two experiments four sheep were fed on dried grass and oat straw in the following combination: (1)straw eaten to appetite and an equal quantity given via a rumen fistula, (2)straw eaten and grass by fistula, (3)grass eaten and straw by fistula, (4)grass eaten and grass by fistula. Mean daily dry-matter intakes (g/kg metabolic live weight, W0·75) for the four treatments respectively were: 13·3, 23·2, 47·8 and 59·4. Dry-matter digestibility coefficients (%) were 37·2, 53·1, 53·3 and 71·3. Changing the digestibility of the diet without changing the component eaten (1 v.2; 3v.4) had a large effect on intake, but changing the component eaten without changing digestibility (2v. 3) caused a twofold change in intake. The latter result implied a difference in palatability between the feeds.
2. The quantities of digesta in the rumen, determined by removal before and after feeding, were generally twice as great for treatments 3 and 4 as for 1 and 2. There were corresponding differences between treatments in the rates of passage of stained particles of feed and of chromic oxide.
3. In the second experiment the treatments were the same except that meadow hay replaced oat straw. Mean daily dry-matter intakes (g/kg W0·75) were 41·7, 55·1, 59·9 and 68·7 for treatment 1 to 4 respectively, and digestibility coefficient were 54·0, 61·3, 62·3 and 72·0. In this experiment there appeared to be only a small difference in the palatability of the feed.