As a result of one detailed experiment and four demonstrations carried out on the University Farm at Cambridge the following conclusions can be drawn:

1. Lucerne drilled at a constant seed rate per acre, in narrow rows (3½ and 7 in.), gave a greater yield per acre than when drilled in rows 10½ and 14 in. in width. Over a four-year period the yield of dry matter per acre from narrow rows was half a ton more than from wide rows. No evidence was obtained that weed infestation is reduced by narrow drilling.

2. The broadcasting of 5 lb. of trefoil per acre at the time the lucerne was drilled had no effect on the total yield or on weed infestation.

3. Wild white clover broadcast at the rate of 2 lb. per acre at the time of drilling the lucerne slightly reduced the yield of lucerne in the first harvest year, but subsequently it served to secure some control of weeds, and thereby favoured the persistence of the lucerne. Differences in yield associated with sowing wild white clover were, however, slight.

4. Commercial cocksfoot broadcast at the rate of 10 lb. per acre at the time of drilling the lucerne led to very considerable increases in total yield (nearly 3 tons per acre of dry matter over a four-year period). The increments were obtained principally during the spring growth, and some loss of lucerne plants was suffered, so that in a very dry summer the total aftermath yield was slightly reduced. Since cocksfoot almost completely controlled weeds and its mixture with lucerne was very suitable for hay or for silage, it is regarded as a plant that may be included in a lucerne ley with advantage.