1. The influence of frequency of cutting on the yield and chemical composition of Pennisetum purpureum Schum under the environmental conditions obtaining in south-western Nigeria was studied in 1953–54. The cutting frequencies adopted being seventeen, eight, six and four times sampling a year, which permitted the plants to grow for intervals of 3, 6, 8 and 12 weeks, respectively, before cutting.

2. When elephant grass was sampled four times a year, at an interval of 12 weeks, a yield of 52·9 tons of green fodder per acre was obtained for the 11-month period (24 June 1953 to 27 May 1954). This yield was 1·8 times more than the green fodder produced from plots sampled seventeen times a year, at 3-weekly intervals, and 1-4 times more than that obtained from each of those sampled six times a year (intervals of 8-weeks) and eight times a year (6-weekly intervals), respectively. These yields, which are probably the highest ever recorded for any grass species grown in unfertilized soils in Nigeria, are also higher than the yields reported under similar conditions and treatments in other tropical countries.

3. The percentage dry matter similarly increased with the length of cutting intervals. Differences in yields of dry matter between the various treatments are more pronounced than similar differences for green fodder yields. The dry-matter yields obtained from either the 6-weekly or the 8-weekly cutting intervals were significantly higher than those obtained from the 3-weekly cut samples, while those from the 12-weekly cut samples were significantly superior to each of the other cutting treatments. The difference in yield between the 8- and the 6-weekly cut samples, however, was not statistically significant.

4. A progressive reduction in yields of dry matter and of green fodder was shown with successive cuttings, particularly in the case of the more frequently cut grasses; the yields obtained during the 1964 seasons were lower than those of 1953. It is possible that the application of fertilizers might arrest this diminution in growth and yield.

5. It was shown that when the grass was allowed to grow for a period longer than 3 weeks, the stems of Pennisetum purpureum Schum tend to mature more rapidly than the leaves. This fact might be responsible for the rapid deterioration in nutrient content, so characteristic of tropical fodders.

6. The percentage crude protein and of silicafree ash fell as cutting intervals increased; the percentage dry matter, crude fibre and nitrogenfree extract rose directly with length of cutting intervals. It was shown that it is best to cut elephant grass every 3 weeks of growth in order to obtain the highest yield of protein and of silicafree ash per acre; this grass should, however, be cut every 12 weeks to obtain the maximum yield of total nutrients and carbohydrates. It does not appear profitable to cut or graze elephant grass at 6- or 8-weekly intervals for either of the above purposes.

7. While green fodder yields were directly, drymatter yields were inversely, related to the degree of monthly precipitation. Generally, periods of high rainfall were followed by high yields of green fodder and low yields of dry matter, those of low rainfall by low green fodder and high dry-matter yields. The percentage crude protein and of silicafree ash of dry matter, fluctuated directly with rainfall; the percentage nitrogen-free extract and of crude fibre showed an inverse relationship with rainfall fluctuations.

It is these latter constituents rather than the former which thus reflect the dry-matter content of the herbage.