It is impracticable to give a detailed summary of the many findings of.this experiment; attention is confined, therefore, to the major aspects and the general principles emerging.

1. By quantitative control of the plane of nutrition, twenty closely inbred pigs have been made to conform to four major variations in the shape of the growth curve from birth to 200 lb. live weight. A high rate throughout (High-High), a high followed by a low rate (High-Low), a low followed by a high rate (Low-High) and a low rate throughout the period (Low-Low) afforded comparison between animals of the same weight but different age and between animals of the same weight and age but with differently shaped growth curves. The relative effects of the treatments upon the development of body proportions and anatomical composition have been studied.

Apparatus for measuring the rate of diffusion of carbon dioxide through granular solids is described and the results obtained with it are shown to conform to the curve connecting D/D0 and S previously obtained for carbon disulphide and acetone. The equation D/D0 = 0·66S, which it is suggested should replace Buckingham's equation D/D0=S2, is applied to a discussion of soil aeration, and it is shown that at all porosities the rate of diffusion of carbon dioxide from the soil is sufficient to account for normal respiration without invoking the assistance of meteorological changes. A further application of the equation to water vapour movement in soils is briefly discussed.

1. An examination of carcass reports from two groups of bacon pigs (comprising eleven litters and seventy-five pigs) was made to determine whether the variation in growth rate of normally reared pigs was associated with differences in carcass measurements. The method adopted was to fit orthogonal polynomials to the third degree to the last 10–11 fortnightly weighings converted to logarithms.

2. The residual standard deviation after fitting was about 2–4%. For present purposes, the fitting of second degree parabolae appeared adequate.

The results are described of twenty-two experiments carried out on the feeding of bacon pigs at different centres throughout Great Britain, in the course of which nearly 400 pigs were fed according to three different scales of rationing. All the pigs were fed on the same scale until reaching 100 lb. live weight, when they were eating 4 lb. of meal per head per day. From this stage three different levels of feeding were followed; in each case the daily allowance reached a maximum when the pigs were 170 lb. live weight. The amounts being fed were then: 7 lb. 8 oz., 6 lb. 5 oz. and 5 lb. 3 oz. per head per day respectively.

1. The influence of storage in the light and in the dark at ordinary temperature and in a heated room at 70–80° F. on the carotene content of finely ground artificially dried grass stored in paper sacks and jute sacks was studied. There was an initial drop in carotene content from 61·1 to 46·5 mg./100 g., i.e. 23·9%, in the first month, and a total loss of 31·4% during 6 months' storage (August to February). No difference could be detected either between treatments or types of containers.

2. There was a marked loss of carotene during 13 months' storage of baled artificially dried grass and of hays, amounting to 30–40% of the original value.

3. Two methods of estimating carotene were compared. The method of Ferguson & Bishop (1936) gave higher results than the method of Peterson, Hughes & Freéman (1937). The difference is probably due to incomplete extraction in the latter method.

4. Chromatographic analyses of “carotene” fractions from the above grasses showed the presence of coloured impurities amounting to 20·5–33·8% of the total recovered pigments.

5. As these impurities are biologically inactive, chromatographic analysis or the use of special solvents are probably necessary for the accurate determination of carotene in forage.

The mineral compositions of the clays from a red earth and a black cotton soil from Hyderabad, Deccan State, India, occurring in close proximity in the field are determined. Both soils are derived from the same or from very similar parent rocks, a coarsely crystalline granite or gneiss.

For both soils there is practically no variation in the mineralogical composition of the clay throughout the profile, but for any given clay there is some variation with grain size. The main contrast between the two is that the red clay contains predominantly kaolinite or halloysite, whereas the black clay contains mainly beidellite, a member of the montmorillonite group. The topography appears to be the principal factor associated with this difference in minerals, and the processes of weathering believed to have produced the contrasted clays are discussed with reference to experiments on the leaching of felspar in the laboratory and on hydrothermal synthesis.

A method is described by which small quantities of indolyl acetic acid can be detected and measured quantitatively, depending on the effect of this substance on osmotic pressure of root hair cells of wheat seedlings. Full details are given of the procedure adopted and the precautions which it was found necessary to take, and suggestions are put forward as to how the method might be rendered more accurate if necessary. Possible sources of e"rror are also discussed.

An experiment is described in which this method is used to determine what effect is produced by different manurial treatments on the content of indolyl acetic acid in the soil. It was found that farmyard manure tended to increase the quantity present.

A new method of analysing the results of experiments arranged in lattice (quasi-Latin) square designs is described. By this method the information contained in the inter-row and inter-column comparisons is recovered. The procedure followed is to estimate the actual accuracy of the inter-row and inter-column comparisons relative to intra-row and column comparisons, and assign weights proportional to these relative accuracies.

The methods of computation by which this procedure can be carried out expeditiously are described in detail, and are illustrated by two experiments carried out at Rothamsted and Woburn. The gains in accuracy obtained in these two experiments (as compared with ordinary randomized blocks) were approximately 30% and 60% respectively, allowing for losses of information due to inaccuracies of weighting.