H.—29.

An investigation was commenced for the Fields Division into the relationship between field responses in L.P.K. manurial trials and the results of soil analysis and analysis of the herbage from the different treatments. The portion dealing with the soils has been published, but the pasture analyses are not yet complete. Fertilizers and Limestones. The usual routine testing of fertilizers and lime supplies and consultation with the respective administrative officers has been continued. Small-scale trials have been commenced of the manufacture of magnesium-silicate superphosphate, using ground serpentine as a reverting-agent with unmatured superphosphate, with a view to reducing phosphatic fixation on certain soil types such as ironstone soils. Waters. Water samples have been analysed in connection with irrigation supplies and reclamation projects, wool-scouring works, liver-fluke investigations, and for stock supplies. Miscellaneous. Special investigations include the identification and determination of phenol in honey for export rejected on account of taint (up to 3-8 p.p.ni. of phenol was found), determination of the amounts of arsenic remaining after several weeks in fleeces from sheep treated with dips for blow-fly control, determination of manganese in citrus leaves from a suspected case of manganese-deficiency chlorosis, and the checking of moisture and methods for its determination in export tobacco. The periodical checking of cattle-dips, meat-marking fluids, &c., has been continued. Supplies of cobalt sulphate and cobaltized salt have been prepared and distributed for sale to farmers. A number of stock-remedies have been analysed and reported on. All meetings of the Standards Institute Chemical Committee were attended. Chemical Control of Ragwort. Mr. F. B. Thompson, Agricultural Chemist at Ruakura, reports as follows : — " At the commencement of the year under review it was known that sodium chlorate and various proprietary chlorate compounds were the most efficient weedkillers for ragwort control, and that small applications would kill the foliage and crowns but only the tops of the roots, while large doses would kill all the roots, the incompletely killed roots being each capable of regenerating a new plant. " The earlier experiments indicated that the Ruakura ragwort nursery was unsuitable owing to the abnormal root-development under cultivated conditions. In the absence of suitable ragwort growing naturally on Ruakura Farm, a comprehensive series of field trials were laid down 011 a farm near Putaruru. These trials were to determine the quantity of chlorate required to produce complete root kills, the best method of application, and optimum seasonal and weather conditions. Treatments were made through 1938 on plots each 16 square yards with a heavy infestation varying from fifteen to thirty ragwort plants per square yard. " Sodium chlorate was applied, both as a spray and as the lime-chlorate dust, at the rate of 37, 75, 150, 300, and 600 lb. per acre, the material being applied evenly all over the plots. Practically all the existing plants were killed, and the efficiency of the treatments was judged by the absence of small plants growing from incompletely killed roots. The results were masked by numerous seedlings from seed already 011 the ground and not destroyed by the chlorate. Counts were made by going over portions of the plots inch by inch inspecting the roots of each plant to distinguish between seedlings and root regrowths. " The results suggest that winter applications are the most effective. This is probably due to the combined effects of frost and sodium chlorate. In February, 1938, during several days of heavy rain, sodium chlorate was sprayed on a series of plots. A month later, during the middle of the drought in March, another series was treated. The counts of root regrowths were— " Wet conditions : 2-2 per square yard (average of twenty-nine). " Dry conditions : 5-3 per square yard (average of nineteen). " Two similar series were treated with chlorate applied as the 5 per cent, lime-chlorate dust. The wet-weather application was made in May, and the dry in February at the beginning of a six weeks' dry spell: — " Wet conditions : o'6 regrowths per square yard (average of nine). " Dry conditions : 4-6 regrowths per square yard. " Plots were sprayed with sodium chlorate using various volumes so that the solution strength varied from 0-5 per cent, to 10 per cent. This was repeated using different quantities per acre. No significant variations were found in the results. " Sodium chlorate was compared with three proprietary weedicides. All four materials gave substantially similar results, whether applied as a spray or the dry dust, under both wet and dry weather conditions. " Applications of 300 lb. or more per acre cause a reduction in the clover and an increase in the Yorkshire fog 011 the plots. " Seedling infestation after chlorate treatment was remarkably heavy. Some of the plots averaged over three hundred plants per square yard, but the usual figure was about thirty. " The Putaruru experiments entailed the examination of the roots of 6,300 plants on a total area of 98 square yards.

70