H—34.

Selection, Testing, and Propagation op Varieties. Only one or two varieties have been added to our collection, but much useful information has been gathered concerning varieties we have had for some time. Strength-tests have been carried out on samples of Ngaro fibre milled from our manurial plots in order to see if any great differences of strength exist among them. Other work has not allowed of extremely detailed work in this connection, but the tests failed to show any significant differences in strength. The general conclusion from these tests of Ngaro is that the manures have made no appreciable difference to fibre percentage or to strength. The Ngaro which was planted in March, 1929, has been ready to cut for a year or so, but cutting was delayed until land was available for extending the area of this variety. Cutting was finally carried out in the last week in February of this year. The ordinary hook-cutting method was employed, but the height of cutting was about 16 in. in order to spare the fans from too severe injury before transplanting. The outside row of plants, having more available root-space, would naturally give a heavier yield. This row was therefore not included in the weighings. The leaf from each of the other plants (originally thirty) in each plot was weighed and recorded separately for each plant. In calculating the average yield per plant in the plots the " neutral" row where two plots met was not counted, nor were plants which were adjacent in the row to plants which had completely died out from yellow leaf. The bushes from which the leaf-samples had been cut for testing in December, 1933, were also omitted from the calculation. Altogether there were twenty-four plots, being four replications of each of the following treatments, the rate of application per acre being given in each case : — (a) Control: No treatment. (b) P : Superphosphate (5 cwt.). (c) P.K. : Superphosphate (5 cwt.) plus 3 cwt. 30 per cent, potash salts. (d) N.P.K. : Superphosphate (5 cwt.) plus 3 cwt. 30 per cent, potash salts. (e) P.N. : Superphosphate (5 cwt.) plus 2 cwt. ammonium sulphate. (/) 8.8. : 8 cwt. blood and bone. The manures were applied on 13th November, 1929, and no subsequent dressings were given. Messrs. Imperial Chemical Industries, Ltd., who donated the ammonium sulphate for this and other tests are specially thanked for these gifts. The general results of the manurial trials in terms of average yield per plant in all four plots subject to each treatment were as follow :— Control .. .. .. - .. .. . . .. 85 • 9 Superphosphate .. .. .. . . . . . . 93 • 6 Superphosphate plus potash . . .. . . . . ..94 • 1 Superphosphate plus potash plus ammonia . . . . . . 92 • 2 Superphosphate plus ammonia . . .. . . . . . . 85 • 2 Blood and bone . . ... .. . . . . . . 94 • 2 Mr. A. W. Hudson has kindly made a statistical examination of the results and finds that on statistical grounds the results do not prove conclusively that any of the manures has increased the yield. In practice it can, however, be said that the increased yields given by four of the treatments as compared with the control stand more than an even chance of being due to the treatments themselves, and not to mere chance. One row running down all twenty-four plots had lime applied for 11 ft. on each side of it at a rate of 2 tons per acre. The average yield per plant in this limed row was 95-8 lb., compared with averages of 88-5 lb., 91-5 lb., and 89-9 lb. for three similar unlimed rows. The manurial trials were set out primarily to investigate the possible effects of manures on the incidence of yellow leaf disease. A study of a plan on which every plant is marked fails to show a beneficial effect in this respect from the application of any of the above dressings. The following facts do, however, emerge from a consideration of this area : — (a) Sudden and fatal attack by the disease frequently occurs in a plant growing otherwise under conditions suitable for vigorous growth. This point is indicated by the extra heavy yield of bushes adjacent (4 ft. away) to the dead plant. (b) In general, the disease occurs sporadically through the plantation, but there is a tendency for two or three adjacent bushes to take the disease in the same year (not in subsequent years). This spread of the disease appears to be along the rows where the plants are closer (4 ft.) rather than between the rows, which are 11 ft. apart. The term " yellow leaf "is not a good description of the disease as seen in this area. The plant frequently shows only traces of yellow colour, but the leaves curl and fall down in a manner which seems to indicate a cessation of water-supply from the roots. The yield of leaf as tested by weighing immediately after cutting was 38-7 tons per acre over an area of nine-twentieths of an acre. The outside row, on account of better opportunities for growth, was not weighed, and in this test the plots which had earlier been side-leafed were not counted. The area considered was one continuous piece of land, not several isolated plots. The area included one-half of the space between these rows and the adjacent ones —i.e., an area 11 ft. by 4 ft. was allowed for each plant originally set out, whether or not these plants had survived. This yield would probably have been increased to 45 tons per acre had the plants been cut at the customary height.

31