special pains are taken to guard against the entry of this impurity into the samples, unreliable results are obtained. Similarly, pumice soils are entirely abnormal in mechanical composition, the particles consisting of an easily broken solidified froth. This cannot be analysed by the standard methods in use for determining the mechanical composition of ordinary soil. If such methods are used they give unreliable results, the finer fractions tending to become higher in quantity owing to the breaking-down of the particles in the course of preparing the sample for the analysis. The clue afforded by the analysis of blood, suggesting the deficiency of iron, has been followed up in the work undertaken by the Department in the Dominion Laboratory and in the field during the past fourteen years. Areas have been leased and field experiments conducted thereon. Finally, a demonstration farm was purchased at Mamaku, in the heart of the bush-sick country, in an endeavour to learn a practicable method of farming this type of land without periodically changing the stock to healthy country, as is the practice at present. This country is more adapted to cattle than to sheep farming, and, could a practicable remedy be discovered, a great area would be available for dairy-farming. The chemical part of the work, which has been under the immediate direction of the writer, comprises the analyses of soils, fodder plants, and animal specimens, and these have not shown the presence of any known poison. The absence of poison is, however, indicated by certain facts. Horses may be kept in perfect health for many years upon the same pasture on which ruminants would die in three to nine months, sheep being most and cattle least susceptible. Ruminants, however, when given turnips and hay made from the bush-sick pasture, can be kept healthy while still grazing on the same pasture which as a sole ration would bring on bush sickness. Molasses, bran, and other imported concentrated foodstuffs added in small quantity to the natural ration enable an animal to be kept healthy on the bush-sickness-inducing pasture which, if it contained a poison, would undoubtedly continue to exert its deleterious effects. Again, when an animal at the onset of the sickness has been sent away for a change to healthy country, and after a few months it returns in poor store condition to the pasture upon which it was becoming sick, it fattens or improves greatly in condition. These facts alone, the writer considers, are sufficient to disprove the possibility of poison being present in the natural fodder—cocksfoot-grass and clovers. The author considers that there can be no other explanation of the cause of bush sickness but that which postulates a deficient food-ingredient. It is not to be thought that the organic nutrients are deficient; grasses and clovers grow particularly well on these pumice lands, and provide an ample organic ration. It must therefore be in the mineral or inorganic portion where one must search for the deficient ingredient of the elements necessary to maintain animal life—calcium, phosphorus, potassium, sodium, magnesium, chlorine, iron, and sulphur, named in the relative order in which they occur in the animal's ashes. Iron is the only element about which there can be any doubt as to its presence in sufficient quantity. Phosphorus, although often deficient in the soil, is obviously not low enough to produce nutrition disease in the animal. Phosphorus is stored in the bone of the animal; bush-sick animals show no disease of the bones or other symptoms usually manifested by deficient phosphorus-supply in the diet. Moreover, administration of phosphates to the animal, either medicinally or through the pasture, does not enable them to be kept permanently free from bush sickness. There may be other elements which