The results of a yield trial may be usefully indicative; they can never be decisive. Geographical differences of soil and climate and seasonal fluctuations, limit the applicability of results to the locality and the year of the experiment. No more can be achieved than to ensure that the data accumulated in the prescribed circumstances have as high as possible a degree- of statistical probability. For this, there must be adequate precautions against a variety of factors predisposing to error. These factors may be broadly classified as follows:

(a) Soil differences within the locality, i.e. major-scale differences as, for example, between the two ends of a 10-acre field.

(b) Soil differences within small portions of the test area, e.g. as between contiguous areas of 1 square ft. or even as between the areas occupied by neighbouring single plants.

(c) Differences in the seeds sown.

(d) Damage by wire-worm, etc., resulting in the killing off of some plants especially in the very early stages of growth.

(e) Irregularities in light intensity and inter-plant competition owing to irregular seed rate or depth of sowing, etc.

(f) Losses during harvesting.

(g) Irregularities in the water content of the harvested material.

1. An apparatus is described which has been designed to impart a known number of vibrations per minute to a suspension of soil or other material.

2. The problem of the calibration of the apparatus is briefly discussed.

3. A new method for the comparison of the mechanical composition of suspensions is described.

4. Various preliminary treatments of samples of soil for mechanical analysis are compared, and it is shown that:

(a) The rate of wetting is an important factor.

(b) A combination of triturition and vibration gives a satisfactory degree of dispersion.

(c) The use of acids is not advisable.

The state of consolidation resulting from the nature and treatment of the previous crop has a most marked effect on the draft, causing variations from 107 to 580 lbs. in gross draft and from 68 to 483 lbs. in nett draft on a typical two-horse soil, the force exerted by each horse being about one hundredweight and a half.

The relation between draft and depth of furrow is linear, the nett draft per square inch tending to increase in value with increasing depth, except when the previous crop was roots, when a decrease at depths greater than 6-8 ins. is noticeable.

The relation between draft and width of furrow is linear, the nett draft per square inch in some cases showing an increase and in others a decrease with increasing width.

Single plot trials owing to the magnitude of the experimental errors, are practically useless as a test for comparing the yields of grain of one variety of oat with that of another. At best the results are only applicable to the particular experiment in question. For the old Scotch varieties of oat the probable error on trials of this kind amounts to about 18 per cent, of the yield of grain and to about 20 per cent, for the new varieties. Adopting single plot trials an 18 per cent, error means that when determining the superiority in the yield of grain of one variety over that of another and when the difference is not likely to be more than about 5 per cent, it would be necessary to have 214 centres with no duplication of plots at any centre in order to endeavour to obtain a conclusive result from a single year's trials. When the difference is likely to be about 10 per cent, it would be necessary to have 53 centres.

Data are supplied showing that the experimental error for different crops also for the grain and straw of the same crop is not the same. It is greatest for the grain and lowest for the hay crop.

In connection with nutritional investigations on poultry being carried out in the above Institute by the writer's colleague, Mr E. T. Halnan, it became necessary to devise a satisfactory method for the estimation of uric acid in poultry excreta. Although numerous exact methods have been elaborated for the determination of the amount of uric acid in mammalian urine, the corresponding problem in connection with the excreta of birds has received relatively little attention. This circumstance is no doubt due in large measure to the peculiar difficulty attaching to the quantitative investigation of the nitrogenous constituents of poultry excreta, arising from the fact that the urine and faeces are not excreted separately. The urine is not temporarily stored in a bladder, as with mammals, but proceeds directly from the kidneys through the ureters to the cloaca and there undergoes admixture with the faeces from the intestines.

The object of this note is to call into question the soundness of the generally accepted method by which the so-called apparent protein digestion coefficients of foodstuffs are corrected.

The digestion coefficient for the protein of a foodstuff as ordinarily determined by animal experiment does not afford a strictly accurate idea of the actual digestibility of that constituent. The figure represents a minimum value, and this arises from the well-known fact that the faeces do not consist solely of undigested food residues, but are to an appreciable extent contaminated by nitrogenous products which have been secreted into the alimentary tract and have escaped reabsorption. The nature of this so-termed metabolic material is twofold, being partly protein (mucus, epithelium, etc.) and partly non-protein (residues of digestive secretions, etc.). Before proceeding to a critical survey of the correction method, which is designed ostensibly to enable allowance to be made for the presence of such products in the faeces, it is of essential interest to consider briefly the true significance of the term “protein digestibility.”

1. Of the three sources of supply of nitrogen, rape cake gave the highest mean yields in the absence of superphosphate; while nitrate of soda gave results significantly better than those obtained from ammonium salts.

2. Superphosphate is of importance to the barley crop, giving greatly increased yield when applied.

3. Sulphate of potash seems to have an adverse influence upon the barley yield.

4. The deterioration of the barley plots is much heavier than on the wheat plots. The convention of ascribing the whole linear term to deterioration may not be as true for some plots as for others. Part of the diminution in mean yield may be due to slow changes other than deterioration.

5. The mean annual percentage diminution is least on those plots in receipt of superphosphate and emphasises the importance of phosphoric acid not only in increasing the mean yield, but in maintaining the fertility of the soil.

6. Barley is more variable than wheat and is more subject to the influence of meteorological conditions.

7. The slow changes other than deterioration are relatively unimportant and seem closely connected with the manurial treatment. The cause of the slow changes is obscure, but there were probably special influences operating on certain of the plots.

It has been shown that maize silage of good quality can be produced with losses of dry matter which are not necessarily much greater than those which characterise the production of oat and tare silage.

Excessive drainage of juice consequent on immaturity of crop contributes materially to the total loss of dry matter.

The main reason for the comparative failure of attempts in this country to produce maize silage successfully from the points of view of quality and economy lies probably in the general use of the late maturing variety, American Horse Tooth. Success will probably depend on growing a variety of maize which is able to reach a desirable stage of maturity under English conditions before being cut for the silo. Preliminary trials have indicated that the necessary qualities may be found in certain varieties like Saltzer's North Dakota, Longfellow, Compton's Early and White Cap, all of which mature at least a month before American Horse Tooth.

Odén has shown that colour intensity comparisons between alkali extracts of peat soils of varying origin can be made since, despite the possible difference in the nature of their humic matter, the absorption spectra are almost identical within the visible range. These absorption spectra are also comparable with that of Merck's Acidum Huminicum, which is a standard preparation. Odén developed his colorimetric method as a means of measuring the degree of humification of the organic matter of peat soils, that is to say, the proportion of the total organic matter which could be classed as humic matter. In this way the disadvantages of Grandeau's method, in which there is no discrimination between the coloured and the colourless parts of the soil organic matter, are obviated.