H.—34.

1. Texture as shown by Mechanical Analyses. (a) Levels Series. —In this series six types can be characterized. Where only a light covering of soil has been deposited, especially on the higher parts, the underlying shingle is well mixed with the topsoil, so that, although the soil between the shingle is still a silt loam in texture, the percentage of stones in the whole sample is sufficiently high to characterize the type as a stony silt loam (14) and to considerably affect irrigation practice. This type is well exemplified on Burns's and Scott's properties. It will be noted that, compared with the typical silt-loam type, both fine gravel and coarse sand are higher in the stony silt loam. In the Levels series, therefore, stony silt loams will be less retentive of water and will allow a faster drainage rate. The most productive type of the Levels series, a deep silt loam (10), is well developed within the scheme, and typical samples can be seen at the Washdyke corner of the irrigation area. The type is very uniform, and only small variations from average figures are found. (b) Kereta Series. —Where the Levels silt-loam type occurs near the coast, the small difference in level between the soil and sea has resulted in a banldng-up of ground-water, so that there is a strip of meadow soils running parallel to the sea-coast. These are still silt loams in texture, and, if they are drained as proposed by Mr. T. G. Beck, there is no reason from a permeability viewpoint why the use of irrigation water should not be as successful as on the same type farther up the plain. No opinion is ventured, however, on the economic aspects of the problem. Table II shows that the subsoils in all cases are silt loams. It can be concluded therefore that all types are suitable for irrigation. (c) Arowhenua Series. —The soils of the Arowhenua series, much more recent in origin than the Levels series, are river deposits. As is to be expected, more variation is encountered, for alluvial soils usually change rapidly in texture. There are nine types in this series, while the Seadown series, with the same parent material, are meadow soils formed in the same way as the Kereta series previously mentioned. The coarser type (5), shown on Kane's and Gaffney's properties, is a fine sandy loam in texture, and what has been said in connection with the stony silt loam can be said here on the relationship of these soils to water. They will not be so retentive of water, drainage will be quicker, and they will need to be irrigated more often with smaller quantities of water to give the optimum results. The main type of this series is again a silt loam (1). The analyses show, however, that the percentage of the coarse and fine sand fractions are somewhat lower in the Levels silt loam, with a corresponding rise in the percentage of the silt fraction. Sample 1175, which is just in the silt-loam class, would probably have been classified as type (8) had a more detailed survey been possible. (d) Seadown Series. —The meadow soils of the silt-loam type show no obvious difference in texture, and the soil should be irrigable since the meadow type is not due to impermeability of subsoil but to a high water-table. (e) Washdyke Series. —ln the Washdyke series two samples have been collected, one more saline than the other. The texture of sample 1001 is a silt loam, but the salinity of the soils would be an effective bar to their agricultural utilization without extensive flooding, drainage, and the use of lime. Table II shows that the textures of subsoils in the Arowhenua series are such as to allow drainage through them. They follow in each case the corresponding topsoil. To sum up, it can be said that the texture of ail soils in the irrigation scheme indicates that they are eminently suitable for irrigation. 2. Nutrient Status. (a) Phosphate. —A study of Table 111 shows that the main point of interest is in the big difference in the available phosphate status between the Levels and Arowhenua series, a difference so marked as to be a more reliable criterion of the Arowhenua and Levels types than any field examination. It will be seen that the older deposits of the Levels series, either because of leaching or of an original lowphosphate status in the parent material, are much lower in phosphate than the more recent riverdeposited Arowhenua series—a dissimilarity apparently reflected in farming results, since the Arowhenua series is reported to be more fertile (Grange). Top-dressing programmes, however, do not indicate that there is a great deal of difference in the phosphate used in the two series, and this fact suggests that top-dressing experiments should be laid, down on the two types to demonstrate the optimum dressings that should be applied. In the Levels series the phosphate level, with one or two exceptions, is definitely low, and increased crops as a result of irrigation practice will necessitate the use of much heavier dressings than have been customary in the past. In the Arowhenua series, with its high phosphate level, there should not be the same necessity for heavy top-dressings, but until fertilizer experiments are laid down it will not be possible to say what rate of top-dressing is most suitable from an economic viewpoint. Certainly the chemical analyses show a high enough level to suggest that no extra phosphate is necessary, but, unfortunately, the chemical method does not always give a clear picture of the soil phosphate readily available, and the use of an easily soluble top-dressing like superphosphate may be beneficial. For instance, in Ashburton, a top-dressing of 1 cwt. superphosphate per acre gives results on the high-phosphate soils of the Ashburton River (Soils of the Ashburton County Chemical Report, Rigg, Kidson, and Hodgson, sth Annual Report of the Soil Survey, 1934-35, p. 27). Similarly A. W. Hudson records (N.Z. Jour. Agric, 1929, 38, p. 338) that on the farm of J. W. Topham, situated on the Arowhenua silt loam, there was no increase from superphosphate one year, but in the next there was a significant increase of seven bushels of wheat per acre. Again, from experiments on the farm of M. J. Fitzgerald also on the Arowhenua silt loam, Hudson and Montgomery (N.Z. Jour. Agric. 1930, 41, p. 350) state that a slight response was visible to superphosphate.

60