After reviewing the possibility of soil deterioration in the Gezira and its dependence on change in the exchangeable bases in the soil we have given an account of field trials with soil improvers. Increased penetration of water, improved supply of nitrogen and higher yields were obtained, but the effects were not lasting. A part of the initial response is ascribed to the flocculating action of salts formed near the soil surface by interaction between soil and soil improver. With continued watering these salts are washed from the surface soil, but the applications of soil improver were insufficiently heavy and the leaching insufficiently thorough to effect any deep-seated change in the soil. Drainage experiments in which effluents were measured and analysed showed small changes in the right direction, but their main outcome has been strikingly to confirm the view that lateral movement of water through Gezira soil is so slight as to promise small hope of achieving any large-scale improvements by means of subsoil drainage. Analyses of saltbush have been made and showed that although these plants remove much sodium from the soil they can hardly be regarded as a practicable agent in soil improvement, since the amount of sodium in the soil is comparatively large. On the other hand, it appears that by including saltbush in rotation and by removing the crop from the land it is possible to guarantee that no progressive deterioration will occur in consequence of accumulation of sodium introduced in the irrigation water. Further analyses of irrigation water were made in this connexion and substantially confirmed Beam's earlier data.

We do not yet know whether, under normal irrigation, the general tendency is for an accumulation of sodium or whether, on the contrary, salts are slowly being washed from the soil column (Greene & Peto, 1934). Until the nature and extent of chemical changes in the soil are accurately assessed it is impossible to say what annual expenditure on soil improvers, drainage, growing and disposal of saltbush is needed for maintenance of soil fertility. Further laboratory work should be directed to the detailed study of base exchange in Gezira soil and to observation of the physical characters on which permeability depends while, in the field, further experiments should be made in the use of soil improvers and drainage; information should also be sought as to the most convenient and least expensive way of including saltbush in the rotations best suited to different parts of the Gezira and as to the economic disposal of saltbush ash.

1. The results of three uniformity trials with a U4 cotton at Barberton are reported. Two of these trials, carried out in a dry year, gave yields in the region of 400 lb. seed cotton per acre; the third, in a wet year, yielded over 1100 lb. per acre.

2. Results from the three experiments agreed closely in essential details regarding size and shape of plots.

3. In all three cases the percentage standard error per plot decreased rapidly as the plot was lengthened, but tended to increase slightly as the width of the plot was increased.

4. The standard error increased as the block size was increased by the addition of more plots, indicating the desirability of keeping the number of strains in a variety trial as low as possible.

5. With plots of the same shape the smallest plots were the most efficient, while with plots of the same size the efficiency increased as the plot shape became longer and narrower.

6. Two fertility contour maps are given, which illustrate a general patchiness in yield common to cotton crops in the Barberton district.

7. Details of the types of plot used at Barberton are given, together with the reasons for their adoption.

Since sugar beet was first introduced into this country many experiments have been conducted to determine the effects of spacing on yield: the general advice based on these experiments is that beet should be spaced as closely as working conditions permit. Davies (1931) carried out a series of experiments in the West Midlands, and came to the conclusion that yield was not related to the number of roots per acre, but was affected by their distribution. His work demonstrated that wide row distances could not be compensated by narrow spacing in the row. He found that yield increased as row distance decreased down to 16 in., but that singling distances of 4—10 in. produced no differences in yield of roots: the yield of green leaves, on the other hand, was increased as singling distance decreased, but was unaffected by row distance (Davies & Dudley, 1929). Although Davies' results would be generally accepted as a true expression of the general rule, many isolated spacing experiments fail to conform to them; it is possible that discrepancies in results, that undoubtedly occur, may be due to variations in the “plant” actually obtained in the experiments. Engledow et al. (1928), as a result of counts and weights taken on ordinary farm crops of sugar beet, concluded that uniformity of “plant” was a most important spacing factor affecting yield. In America Brewbaker & Deming (1935) have found yield to be related to percentage stand (correlation coefficients varying from +0·35 to +0·70), the regression between the two variables being approximately linear over the range studied. They also found that uniformity of “plant” was more important than spacing distances, either between or in the rows. Their work showed that single gaps had little effect on yield, because neighbouring beet compensated for them to the extent of 96·2%; serious loss of yield only occurred, therefore, with adjacent gaps. Pedersen (1933) studied the relationship between percentage of gaps and yield in a large number of Danish experiments with sugar beet and mangolds. In the case of sugar beet he found that the compensatory growth of neighbouring roots amounted to 76% for a single gap, and that the percentage compensation decreased as the size of gap (i.e. number of missing beet) increased. In an earlier paper (Pedersen, 1931) he had shown that under ordinary field conditions the distribution of gaps was approximately random.

1. Yields of plots estimated from the known weights of “perfect” beet (i.e. plants with no gaps in the ring immediately surrounding them) showed 18 × 9 in. to be the optimum spacing.

2. Sugar analyses performed on individual “perfect” beet showed that very diminutive roots were low in sugar, but that in general sugar percentage decreased with increasing weight of root; the decrease was, however, slight and only amounted to unity for an increase of 1000 g. in weight of root.

3. In a dry year the roots immediately surrounding a gap compensated to the extent of 80–89% for the missing plant: the allocation to the individual neighbours was approximately inversely proportional to the square of their distance from the site of the gap. In a wet year compensation was less complete, amounting to from 41 to 84% under various spacing treatments. In both years compensation was less complete in the case of tops than in the case of roots.

4. Within classes of beet similar in regard to spacing treatment and gappiness there still remained considerable variation due to soil and to genetic heterogeneity.

5. The genetic variability of commercial seed appears to be large, so that in sugar beet experiments it is very desirable that each plot should carry a considerable number of plants: in these experiments, assuming that all variations within ultimate classes were of genetic origin, 400 plants would have been necessary to reduce the plot error due to genetic variability to 2% of the mean.

1. The distribution of the total weight of beet in small areas (2 yards by 1 yard) was examined and estimates made of the effects of missing beet on the final yield, and of the yield to be obtained with various percentage plants.

2. The distribution of gaps over the field is found to be non-random. The percentage of gaps varies from block to block, but does not differ greatly between the spacings. Three methods were employed and gave consistent results.

The design and field technique of two large scale experiments, laid down to investigate the problems of sampling and “propinquity”, are described in detail. These experiments were designed so that the plant number, stem number, ear number, straw weight and grain weight for 7200 lengths of 6 in. of drill row, together with the position of each observation, might be obtained.

The lowest sampling error, expressed as a percentage of the mean, is obtained by using the smallest sampling unit, but the large number of sampling units of this size that would have to be taken make it impracticable. The optimum sampling unit consists of 6 ft. of drill row, taken as 3 ft. of two adjacent rows. The size of the sampling unit is of greater importance than the shape in determining its accuracy, and little s lost by using sampling units of 18 in. of five adjacent rows.

The nature of the observation affects the degree of sampling. In the experiments examined the plant number, stem number and ear number require very similar sampling, but grain weight requires a sample which is about twice as large as that required by the other observations.

The percentage of the plot that must be taken as a sample (the sampling percentage) to obtain any given accuracy is related to the size of plot: the larger the plot, the lower the sampling percentage necessary. To obtain an error of 5% of the mean the sampling percentage for grain weight is approximately 5% in 1/20 acre plots, 15% in 1/100 acre plots and 43% in 1/600 acre plots if the plots are not subdivided. These percentages are similar to those calculated from various randomized block experiments in Cambridge.

Analysing Hammond's data on the growth of limbs in the sheep, Huxley (1932) found a post-natal simple negative heterogony, or allometry, as in the latest terminology (Teissier, 1937), of the whole limb relative to the body. As lambs, and the young of most other grazing mammals, are born with relatively long limbs enabling them to follow their dams (Hammond, 1932), the same author pointed out that the leg must have exhibited positive allometry during foetal development.

Having possession of a sufficient body of measurements on the skeletal development of female calves of the Chianina breed, a very tall beef and draft breed of cattle of Central Italy, I investigated the course of growth of height at withers relative to the growth of the length of the trunk (from shoulder to ischium). The growth of height at withers can be considered as a sufficiently reliable measure of the growth of forelimb length owing to the fact that the two principal angles in the limb (scapula-humerus and humerus-radius) do not change very much with age (Engeler, 1935).

It was found that the growth of forelimb length in this breed of cattle exhibits a simple negative allometry relative to the growth of trunk length. This simple allometry, with growth constant about 0·78, holds from birth until over the age of 3 years, that is, throughout practically the whole post-natal growth.

Lack of data after this age leaves open to further research the course in the ultimate period of development.

I thought it would be of some interest to compare various breeds of cattle differing rather widely in the absolute values, as well as in the proportions of height at withers and trunk length at birth and in later life.

The influence of three feeding treatments of differing protein content on growth and carcass quality in bacon pigs has been investigated by the individual-feeding and group-feeding techniques. The protein supplement, which was used in conjunction with barley meal, weatings and a small allowance of lucerne meal, was composed of a mixture of ex. soya-bean meal, dried separated milk and feeding blood meal. Expressing, on the nitrogen basis, the amount of protein food in terms of its equivalent of white fish meal, the standard-protein treatment B supplied the equivalent of 9% of fish meal up to 90 lb. live weight, 7·6% from 90 to 150 lb. live weight and 3·8% from 150 to 200 lb. live weight. The corresponding figures for the low-protein treatment A were 4·5, 3·8 and 1·9%, and for the high-protein treatment C, 18, 15·2 and 7·6%.

Considering the period of feeding up to 200 lb. live weight as a whole, it was found that differences of feeding treatment gave rise to no significant differences in respect of mean rate of live-weight increase and efficiency of food conversion (see Table VI). Only in the earliest stage of the feeding period did the pigs on the low-protein treatment A show a slightly, though significantly, lower rate of live-weight increase and poorer efficiency of food conversion than the pigs on treatments B and C, but such differences had ceased to be manifested by the time the pigs had arrived at 60 lb. live weight, and the slight initial disadvantage experienced by the low-protein pigs was wiped out during the later stages of the feeding period.

1. The quality status of Danish and English Wiltshire bacon sides representative of the highest grades produced at the present time has been investigated by the use of carcass measurements. The data resulting provide “standards” for practical guidance in stock improvement work on bacon pigs, and for the evaluation of experimental treatments involving bacon quality.

2. The variability of the respective characters has been compared by means of the coefficient of variation of each. The results indicate that while selection on a basis of “external characters” and of such internal characters, as are readily measurable, leads to marked uniformity in respect of these characters, it does not necessarily involve similar uniformity in respect of important “internal characters” which are not taken into account in either stock selection or commercial grading practices.

3. External characters thus do not provide a reliable indication of the internal quality status of the bacon pig, and since it is the latter which largely governs the cutting value of the side, concentration upon addition to the former is desirable if still further improvement in the quality of the bacon pig is desired.

4. Variability appears to be affected also by the rate of development of the character concerned; late developing characters in general tend to be more variable than early.

5. The intimate association between nutrition and the rate of development of the various parts and tissues of the pig provides a method of control in addition to that of selection.

Six litters of pigs farrowed, reared and fattened on a ration of barley meal, weatings, soya-bean meal, meat meal and minerals failed to thrive and in the later stages of fattening many of the pigs showed typical symptoms of vitamin A deficiency.

Six comparable litters raised under similar conditions and on the same meal mixture with ° 1 or 2% of a commercial cod-liver oil of guaranteed purity and standardized vitamin content made good progress and nearly all the pigs weaned in these litters were fattened for pork or bacon.

Estimations of liver reserves of vitamin A of the pigs receiving no cod-liver oil indicated that the liver store was exhausted soon after weaning. All pigs receiving cod-liver oil gradually increased their liver store of vitamin A. The stores rose approximately in proportion to the level fed and period of feeding. There was no evidence that the higher levels of cod-liver oil had any practical advantage over °%.

The symptoms shown by the pigs receiving no cod-liver oil included loss of appetite, cessation of growth, impairment of vision in daylight, abnormal gait, convulsive fits and nervous collapse. Pneumonia and/or inflammation of the intestines was found in all of seven pigs that died during the fattening period.

1. Two experiments have been carried out on the Cambridge University Farm to test the desirability of including late-flowering red clover in the seeds mixture in the case of a 4-year ley, and to compare grazing with cutting for hay in the first harvest year.

2. The majority of the red clover survived for the first 2 years, during which it increased the yield of dry matter very markedly; it did not lead to any reduction of yield in the last 2 years. By producing an early cover of the ground red clover checked the encroachment of weeds; although it reduced the proportion of wild white clover in the herbage in the later years, the reduction was not serious.

3. Where the young plants were well established at the time, and normal weather conditions were experienced, grazing was definitely preferable to cutting for hay during the first harvest year. Where the plants were slow in establishing themselves, and in a superlatively dry spring and summer, grazing checked the development of the sward too much, and cutting for hay gave much better results.