Results are reported of sixteen replicated experiments on spring-sown barley and oats in south-east England to compare fertilizer sown in the drills with fertilizer broadcast. In thirteen of the experiments, fertilizer in the drills gave significantly greater yields of grain and straw than fertilizer broadcast, in two of the experiments the differences between the two were not significant; in only one experiment was the grain yield significantly greater where the fertilizer was broadcast.

Fertilizers were twice as effective when sown in the drills as when broadcast. On an average, broadcast fertilizer increased the yields of grain and straw by 4·0 and 6·3 cwt. per acre respectively. Drilling the fertilizer with the seed gave over double these increases, viz. 8·0 cwt. grain and 14·6 cwt. straw per acre.

Drilling the fertilizer with the seed tended to delay brairding. The crop on the fertilizer-drilled plots soon caught up and overtook that on the fertilizer-broadcast plots. Drilling the fertilizer with the seed increased tillering and decreased weed-growth.

Not more than about 3 cwt. of a soluble fertilizer per acre should be sown in the drills or the stand of plant may be poor. For the same amounts of soluble plant food, the higher the analysis of the fertilizer the less is the delay in brairding.

The possibility of obviating the delay in brairding by sowing the fertilizer near, but not in contact with, the seed is briefly considered.