Nineteen experiments were made between 1964 and 1967 on fields where previous sugar beet crops showed symptoms of magnesium deficiency. None, 2·5 or 5 cwt/acre kieserite or 20 cwt/acre dolomitic limestone were tested in a factorial design with none or 3 cwt/acre agricultural salt (crude sodium chloride), and 0.8 or 1.2 cwt/acre nitrogen as ‘Nitro-Chalk’. Additional plots tested kainit (7 cwt/acre) and a large dressing of potash (2 cwt/acre) as muriate of potash.

Kieserite and dolomitic limestone increased sugar yield and the most effective dressing was 5 cwt/acre kieserite, which gave 3·1 cwt/acre more sugar than the crop without magnesium fertilizer. Agricultural salt and the larger dressing of nitrogen were profitable, and neither interacted with magnesium on average; the large dressing of potash also increased yield. The magnesium in the kainit increased yield slightly, but the dressing tested supplied too little to satisfy the crop's requirement of magnesium.

Each year in late summer the percentage of plants showing magnesium-deficiency symptoms was recorded, and a sample of twenty-four plants harvested from each of the magnesium treatments and analysed. All the magnesium fertilizers increased the concentration of magnesium in leaves, petioles and roots, and also decreased the number of plants showing deficiency symptoms.

The magnesium concentrations in plants grown without magnesium differed widely and were related both to the yield response to magnesium fertilizer and to the percentage of plants with deficiency symptoms. Both relationships showed a similar ‘transition zone’ from deficiency to adequate supply, for leaves this was 0·2–0·4% Mg, for petioles 0·1–0·2 Mg and for roots 0·075–0·125 % Mg in the dry matter.