H.—34

plants which are eaten by stock in areas affected with bush sickness. In addition, the effect of fertilizers, including iron compounds, on the intake of iron by different plants has been tested in a series of pot experiments conducted under controlled moisture conditions. In the field further animal tests have been conducted at Glenhope—a typical bash-sick area of the Waimea County, Nelson—in order to ascertain the relative efficiency of soil, limonite ore, and ferric ammonium citrate in controlling bush sickness. Another important line of investigation is connected with the potential hydrocyanic acid content of different pasture plants in the Nelson District. This work has involved the exploration of laboratory technique in connection with the estimation of potential hydrocyanic acid production in white-clover samples. Our field experiments relating to the effect of nitrogenous manures on the yield and chemical composition of typical Nelson pastures have been continued, and some of the more important features emerging from these investigations are summarized in this report. I. Investigations relating to Bush Sickness. (a) Available Iron Content of Soils. In a previous report mention has been made of a large number of determinations conducted by Miss E. B. Kidson in connection with the available iron content of volcanic soils in the North Island. For the extraction of the iron from the soils N/10 oxalic acid was employed. The results obtained on the volcanic soils showed that samples collected from bush-sick localities, including that of the Glenhope district, Nelson (non-volcanic origin), gave an average figure for available iron of 0-57 per cent. Fe 2 0 3 . The limits of available iron in these nine samples were 0-43-0-69 per cent. On these soils stock ailment had definitely occurred, and in certain cases serious mortality had been experienced. Volcanic soils not definitely associated with bush sickness contained much larger amounts of iron soluble in N/10 oxalic acid. The examination of the " available " iron content of soils has been extended during the period under review to sedimentary and other non-volcanic soils. Soil-samples representing twenty-seven distinct types, mainly collected in the Nelson Province, have been, examined for iron content. As a rule, comparatively high figures for iron were found, in certain cases attaining 6-9 per cent. of iron calculated as Fe 2 0 3 . Two outstanding exceptions, however, in regard to high iron content were found in the gum-land soils of North Auckland and also in the pakihi soils of the West Coast and Takaka. In these two cases the percentage of iroia soluble in N/10 oxalic acid was considerably lower than the lowest figure found on recognized bush-sick soils. Although the area of established pasture on these two soil-types is not extensive, bush sickness so far has not been reported in either case. In the experiments of the Cawthrpn Institute conducted on reclaimed pakihi land at Westport, sheep and dairy cows have been grazed on established pastures without detrimental effects, and no symptoms similar to bush sickness have so far been observed. Although, therefore, a low content of available iron has been found in every soil subject to bush sickness, it must not be assumed that a low content of available iron in the soil is, in itself, responsible for the appearance of bush sickness. (b) Iron Content of Plants. A wide range of pasture, native plants, and of fodder crops has been examined by Dr. Askew for iron content. In certain areas affected with bush sickness in the North Island farmers consider that the eating of certain native plants by stock overcomes or reduces the incidence of bush sickness. As it seemed possible that trees and shrubs, owing to their deeper-rooting habit v might obtain iron from substrata not available to pasture plants, a number of samples of green leaves from native plants eaten by stoc"k was analysed for iron content. The plants included mahoe, fivefinger, hangehange, monoao, Carmichaelia flagelliformis, and Cladium teretifolium. With the exception of Carmichaelia, the iron percentage of these native plants did not rise above 0-010 per cent. The majority of the plants ranged about 0-008 per cent, for iron, although in certain cases figures as low as 0-004 per cent, were found. The data which have been obtained for iron on these native plants do not indicate that the animals grazing in the bush receive in their diet a greater amount of iron than if their grazing was restricted to typical bush-sick pastures. In view of the fact that various observers have stated that stock fed on turnips in bush-sick localities show an improvement in health, samples of turnips grown in different parts of the Nelson territory have been analysed in order to ascertain their iron content in comparison with pasture plants. In the case of five samples of Imperial Green Globe turnips collected from different soil-types in the Nelson District very little difference was found in the iron content. The percentage of iron in the leaves ranged around 0-01 per cent. Fe and in the roots about 0-0045 per cent. Fe. A sample of swedes grown on bush-sick granite soil at Glenhope contained a lower percentage of iron than that of the turnips previously discussed. It is impossible to state, owing to the fact that no other samples of swedes were analysed, whether this lower content of iron in the Glenhope swedes is a varietal effect or is due to a lower iron content of the soil. In view of the small variation found in the iron content of turnips grown on different soil-types it would seem probable that the lower iron content of the swedes is more a varietal effect than one due to soil properties. The results of the iron analyses of the turnip samples do not suggest, however, that turnips provide a greater supply of iron" than that contained in typical bush-sick pasture.

20