Kinetic studies of thiaminase I in extracts of ruminant faeces showed that the affinity for one substrate varied with the concentration of the other substrate in the manner of a two-step transfer mechanism. When the alternate substrate concentration was optimal, the apparent Michaelis constant (Km) for thiamine was 176 μΜ and the apparent Km for aniline was 3·19 mΜ. It is recommended that in routine thiaminase assays, the thiamine and aniline concentrations should be at least 1·5 and 25 mΜ, respectively. When non-saturating concentrations of thiamine are used in thiaminase assays the results should be reported as unimolecular reaction constants since the enzyme activity can be calculated only if suitable Km data have been determined.

Improved radioactive and colorimetric thiaminase assays with saturating substrate concentrations gave similar results. The analytical variation of the colorimetric method was rather high but this method may be useful for laboratories which lack radioactive isotope facilities.

Thiaminase assays were performed on cultures of 14 species of rumen bacteria. Only Megasphaera elsdenii had thiaminase activity and its cosubstrate specificity was different from the rumen thiaminase associated with cerebrocortical necrosis in ruminants. It was concluded that the source of rumen thiaminase in that disease has yet to be identified.

Magnesium deficiency has been recognized to occur on acid (pH 5·5 or lower), sandy and highly leached soils or in soils whioh have received heavy application of fertilizers especially with materials which are lacking in magnesium or high in potassium (Berger, 1962; Foy & Barber, 1958). In addition to absolute deficiencies, induced forms can also occur, especially when K:Mg ratio in the soil exceeds 5:1 (Caldwell, Hooper & Little, 1968).