1. Deep tillage, namely, ploughing to a depth exceeding 12 in., or subsoiling to a depth of about 18 in., increased crop yields on about half the fields in which an experiment was made. Unfortunately, it was not possible to recognize what was the difference in soil properties between the 50% of fields that responded appreciably to deep tillage and the remainder whose yields were either unaffected or sometimes reduced by deep tillage, though, as one would expect, sands were normally less responsive to deep tillage than the heavier soils.

2. On the clay soils, deep ploughing in autumn tended to give a surface soil that dried out up to several weeks earlier in the spring than land that was shallow ploughed. Subsoiling did not cause the surface soil to dry out in the same way. Only rarely did the subsoil brought up by deep ploughing give difficulty in working down to a seedbed, and probably in as high a proportion of fields the subsoil worked easier than the surface soil.

3. On the clays and loams, the response of potatoes to deep tillage was more marked on fields that were heavily manured than on those lightly manured.

4. Sugar beet normally gave a higher yield if its potash and phosphate was spread on the land before ploughing, even if this was in the autumn, than if it was applied in the seedbed. The exceptions to this result were for soils heavier than sandy loams if ploughing depths exceeding 12 in. were used, for the yield of beet was, on the average, depressed on these soils if the potash and phosphate was ploughed in compared with it being worked into the seedbed.

5. Deep ploughing tended to give better control of many perennial weeds, and often of annual weeds than shallow ploughing. It is suggested that much of this benefit could be obtained when ploughing to 9–10 in. deep if ploughs were used that cut a furrow considerably wider than the conventional 10½–11 in. furrow, so allowing more complete inversion of the furrow slice.

Results are recorded of experiments over four years to study the effects of various rates of N fertilizer applied early (soon after each cut) and late (1 or 2 weeks before cutting) to grass-clover swards cut repeatedly for conservation.

N fertilizer proved essential for high yields of dry matter and protein per acre. It reduced the amount of clover, but it is clear that clovers are not as productive as grasses adequately manured, and that any N excreted by clovers into the soil is far from enough to enable associated grasses to make full growth.

The amount of N fertilizer applied proved more important than the time of its application. Early N gave higher yields of dry matter but rather lower percentages of crude protein in the herbage than late N dressings.

To increase yields substantially and at the same time to maintain the protein content of the herbage on a par with a grass-clover ley it seems necessary to apply a minimum of 4 cwt. ‘Nitro -Chalk’/acre as an early dressing before each cut. An additional 2 or 3 cwt. ‘Nitro-Chalk’/acre may be applied from one to two weeks before cutting if even higher crude protein values are required.

The non-protein nitrogenous (NPN) fraction of grassland herbage given a range of repeated nitrogenous fertilizer treatments has been examined. The herbage was intended for conservation and the treatments were 0, 2, 4, and 6 cwt. ‘Nitro-Chalk’ per acre per cut, applied early, after cutting, and late, 10–14 days before cutting.

1. Two hundred and ninety-one samples of semen from seventy-four dairy bulls in use at seven artificial insemination centres and one bull-rearing centre were tested and graded according to the type of metabolic response under the experimental conditions described by Melrose & Terner (1953).

2. Conception rates of bulls were obtained from the breeding records based on not less than 150 first inseminations. Since the mean conception rates of centres varied, the fertility of each bull was assessed in relation to the mean conception rate of the centre at which the bull was standing. Bulls were considered of ‘high’ fertility when their conception rates were above, and of ‘low’ fertility when their conception rates were below a dividing line 5% below the mean conception rate of the centre.

3. A highly significant relationship (P < 0·001) was found between the groups of the metabolic responses and the fertility levels of the bulls. The practical application of the test in the assessment of the potential fertility of individual bulls is discussed.

The rate at which organic matter added to soils decomposes under temperate conditions seems to have been little studied. It depends, as Russell (1950) remarks, on the soil aeration, the calcium supply, and the temperature, and tends, in an arable soil, to an equilibrium value depending on the crop rotation practised for a given soil under given climatic conditions. And while this equilibrium value is lowest for a rotation containing a high proportion of wide-spaced intertilled crops or cultivated fallows, it increases with crops of small grains or grass. The results at Rothamsted, Broadbalk Field, as judged by the nitrogen content, show that this equilibrium value with no organic additions has remained almost constant from 1865 to 1945 and similarly, at a higher level, where 14 tons of farmyard manure are applied annually, has remained constant from 1914 to 1945.

The existence of a high degree of variability in all production traits of dairy cattle in tropical countries has been demonstrated. Evidence has been adduced to show that this is due largely to increased environmental variability than to any great inherent differences between animals. The conclusion is drawn that, despite minor differences, the genetics of milk production in tropical dairy cattle is fundamentally very similar to that in European stock.

Statistical analysis of data from calcium and hosphorus balance experiments with laying hens has shown that, for regression equations relating shell calcium to calcium retention:

(a) Individual birds give different regression coefficients (b values).

(b) The equations obtained for each of twelve birds belong to one family which can be represented by a more general equation.

(c) Over the whole of an experiment the b value and the phosphorus balance for each bird is significantly correlated with the total shell calcium secreted by that bird.

The metabolic significance of these findings is discussed in relation to shell formation and bone mobilization, and it is suggested that potentially high secretors of shell calcium are those which have a greater ability to lay down calcium on the bone during periods of low production and a greater ability to remove it from the bones when in high production.

1. When Cheviot ewes were fed a daily ration containing about 5 g. of calcium during pregnancy and lactation there was a loss of 6·5% of skeletal mineral by mid-lactation. Two months after the end of lactation this loss was fully replaced.

2. When the daily calcium allowance was about 2 g. the loss was 18·2% by mid-lactation and was not replaced by 2 months after the end of lactation. However, if the calcium intake was raised to about 5 g. at mid-lactation the skeleton was fully recovered by 2 months after the end of lactation.

3. Resorption was always more marked in certain bones than others, the cervical vertebrae, skull and mandible being more sensitive than the ends of the radius and metacarpal and these ends more sensitive than the shafts. This follows the pattern of resorption reported in Part I of this series.

4. In sheep on low calcium intakes the serum calcium levels rose in early lactation and then fell. If extra calcium was given from mid-lactation onwards serum calcium levels rose to normal. If extra calcium was not given the fall in serum calcium levels continued until the end of lactation, when they rose. Serum phosphate levels had a strong tendency to vary inversely with serum calcium levels.

5. The differences in the mineral status of the skeletons, provoked by differences in dietary provision of calcium of the order used in this experiment, were readily seen in radiographs of flesh-free bone. Differences of this magnitude could not be accurately identified in radiographs of live animals.

1. The results of a survey of the methods of milk production on a random sample of farms in four counties of England and Wales were used in conjunction with bacteriological and keeping quality tests of the milk produced, to estimate the influence on milk quality of factors in equipment and technique.

2. Of the many factors in technique investigated, only the sterilization of utensils and the cooling of the milk appeared to affect milk quality. The effects of these two factors differed in some degree with the four tests used.

3. Milk samples from machine-milked herds had significantly greater thermoduric colony counts than samples from hand-milked herds. No other factor of equipment showed appreciable association with milk-quality tests.

4. Brief recommendations are made about the conduct of any future field survey of factors affecting bacteriological and keeping quality of milk.

1. Methods of estimating water retention in the body of cattle are discussed, and it is shown that methods based on determinations of the initial and final body-water content are of low accuracy and are not suitable for the indirect estimation of fat balance over a period of 24 days.

2. Analyses of tissues from young and adult cattle showed that with the exception of serum, brain, skin and bone, water content of the tissues may be predicted from the following equation:

Water(g./100g.) = 0·292 Na(mg./100g.) + 0·147 K(mg./100g.).

This equation shows that 1 m.equiv. of the Na of tissues is associated with more water than is 1 m.equiv. of the K.

3. No differences between cattle aged from less than 1 week to more than 5 years were observed in the relationship between water and Na and K content.

4. The equation applied also to the contents of the digestive tract, other than those of the abomasum which gave low results, ascribed to the presence of a high concentration of H+ ions.

5. Analyses of whole foetuses for Na, K and H2O showed that their water content could be predicted accurately. The same was true of amniotic fluid, but the water content of allantoic fluid was underestimated by 50%.

6. It was shown that despite the anomalous behaviour of brain, serum, skin and bone, accurate predictions of the water content of the animal could be made from simultaneous determinations of Na and K retentions, providing a correction was made for the storage of bound Na in bone.

7. Data are presented which indicate that the equations apply equally well to the tissues of the sheep.

1. A series of simultaneous determinations of the Na, K, N, C, ash, H2O and energy retentions of calves have been made in order to test the validity of a method of estimating energy retention without recourse to respiration chamber technique. The experiments involved 72 days' calorimetric observation with six animals. A further series of fourteen experiments with sheep involving 58 days' calorimetric observation were made but are not reported in detail

2. It was found that, with calves, water retention could be estimated by an indirect method from Na, K and ash metabolism without significant systematic error. The standard deviations attached to the indirect estimates appeared to be invariant with length of period. Thus, by increasing the length of the period the accuracy of the mean daily estimate of water retention could be considerably increased.

3. With calves, energy retention measured indirectly from body-weight gain and the gain of ash, protein and water did not differ significantly from that determined calorimetrically by the C and N balance technique. The standard deviation of the mean daily difference fell with increasing length of period, being ± 407 Cal./24 hr. in an experiment lasting 8 days. It was calculated that experiments with four calves lasting 24 days would give results of the same precision as that obtained in a single calorimetric trial.

4. The use of the ratio calorie retention/protein retention could permit the calculation of energy retention from the N metabolism of the animal. It is suggested that this method, if developed to take into account the effect of nutritional plane on the ratio, could have considerable practical value.

5. It was found that cutaneous losses of K by the sheep in its skin secretions (suint) make it impossible to interpret the difference between the intake and faecal and urinary losses of K as a body gain of K and express it in terms of water. The method is not therefore suitable for the sheep.

6. The difficulties involved in the application of the method to adult cattle are pointed out, particularly those that stem from the high K content of most of the foods they eat. A very high analytical accuracy is necessary if the estimates of Na and K retention are to have the necessary precision.

1. The pigs in the three experiments reported were weaned at 10 days old and fed water and drymeal diets ad libitum until they reached 25 lb. live weight. They were housed individually in cages and records were kept of each pig's food consumption and of its live weight every third day.

2. Most of the diets fed were based on roller-dried skim milk, rolled oat groats, white fishmeal, sucrose, dried yeast, cod-liver oil, minerals and antibiotic.

3. Forty-eight pigs were used in Exp. 1, which was designed as a 4 × 2 factorial. Food-conversion efficiency was significantly improved by 10% when the crude protein in the diet was raised from 24 to 29%. Increasing the protein to 34 or 39% brought no further significant improvement. There were no significant differences between growth rates which were attributable to the four levels of protein.

The inclusion of 10% sucrose in the diet did not result in increased food consumption, but there were significant improvements of 8% in growth rate and 10% in food conversion efficiency. There were no treatment interactions.

4. Twenty-four pigs were used in Exp. 2, which was designed as a 2 × 2 factorial. Figs fed a diet containing 7% fat grew 4% faster and 6% more efficiently than others fed a diet containing 3·7% fat, but neither difference approached statistical significance. There was no advantage in feeding the meal as a slop, indeed, there was a tendency for the pigs to scour more than on dry feeding. As in Exp. 1 there was no treatment interaction.

5. Forty-eight pigs were used in Exp. 3, which was designed as a 2 × 2 × 2 factorial. The basal diet contained about 30% crude protein and 10% sucrose.

An increase in the antibiotic level from 18 to 112 mg./lb. diet significantly increased growth rate by 7%, but the 9% increase in efficiency of food conversion was not statistically significant.

The addition of a trace mineral supplement significantly improved rate of gain by 7%, but the 6% increase in efficiency of food conversion was not statistically significant.

The addition of a complex vitamin mixture had no significant effect upon either rate or efficiency of growth.

6. In all three experiments the shape of the growth curve of pigs weaned at 10 days old wag divided into two periods. During the initial ‘check period’, lasting about 10 days, the pigs ate only small amounts of food, grew very slowly, and sometimes scoured. After the ‘check period’ there was a sudden change to the ‘growing period’, during which the pigs ate increasing amounts of meal and grew rapidly.

The only treatment which significantly affected the length of the check period was in Exp. 3, when the addition of the vitamin supplement to a diet containing a high antibiotic level shortened the check period by 3 days but then counteracted this advantage by reducing rate of gain during the growing period.

7. Between 25 lb. weight and 8 weeks of age twenty-four pigs from the experiments reported were fed ad libitum on a dry-meal diet containing 22% crude protein. The average live weight of these pigs at 8 weeks was 49 lb.

1. The differences between strains in respect of tripping, pollination, self- and cross-fertilization, the number of pods and seeds set under comparable conditions, and the influence of weather and insects on these aspects was investigated during 1953–5 at Cambridge. Early, mid-season and late strains of lucerne were included in the experiments.

2. Early strains under natural conditions had a significantly higher percentage of florets tripped, and pods formed, than mid-season or late strains. The ease with which early strains tripped, and the likelihood that biological characters of their flowers were better adapted to environmental conditions, are possible reasons for their high percentage of pods set.

3. Tripping is essential for seed setting; at least 98·5% of pods resulted from florets tripped by insects or automatically.

4. There was little difference between strains in the percentage of self-fertilized florets, although in the early strains there was a slightly higher percentage, as well as a higher number of selffertilized plants. There was little difference between strains in the average number of seeds set per pod which was about 1·9 for artificial tripping and selfing and 1·5 for selfing under continuous cover. Under natural conditions, the estimated proportion of pods formed due to selfing was 20% in Eynsford and Du Puits, 12% in Provence and 16% in Grimm. The difference between strains in the percentage of pods formed by artificial crossing was small. The average number of seeds set per pod was 3·9 under natural conditions and 6·7 under artificial crossing for all strains.

5. Sunshine and high temperature encouraged tripping while rainfall suppressed tripping and pod setting. When the effect of temperature and rainfall were eliminated, duration of sunshine was the most important factor influencing tripping. Some variations in tripping and pod setting were found among plants of all strains in different seasons.

6. The population of wild bees was very low. Honeybees and other small insects caused some accidental tripping and crossing, especially in early strains.

1. The weights at birth of 822 pairs of twin lambs have been studied with respect to the differences between males and females. Data were obtained during five lambing seasons from various breeds and crosses in five flocks.

2. The sex distribution of twin pairs came close to the ratio .

3. Analyses made within season and within the major breed categories showed that sex differences were small relative to other sources of variation, but that there was a significant effect of sex of co-twin on birth weight. The following average differences were found (the sex of co-twin being indicated in brackets):

These differences had an average S.E. of 0·078 lb. (Table 3). The sex difference in litters of mixed sex was twice as large as that obtained from like-sexed litters.

4. The data for particular flocks and seasons suggest that in addition to these average effects, there may be some interaction between sex and breed or season which can cause deviations from the averages.

5. These results are interpreted to mean that there is competition between twins for some element of growth in limited supply; and that in twins of mixed sex, the result tends to favour the male.

1. Selections from a cross of varieties of Triticum vulgare and T. Turgidum have been backcrossed repeatedly to the turgidum and vulgare parents and selections of turgidum and vulgare type respectively made within each series.

2. The yield and morphological stability of turgidum- and vulgare-type selections from the straight cross have been compared with those of selections backcrossed up to four times to T. turgidum and T. vulgare respectively.

3. In the case of backcrosses to T. Turgidum the highest yielding selections have been derived from the most frequently backcrossed material, but the vulgare-type selections from the straight cross have outyielded those from the material backcrossed to T. vulgare.

4. An analysis of the yield components of the parents and of the most promising selections, in terms of ears per plant, spikelets per ear, grains per spikelet and 1000-grain weight, has shown no evidence of a combination of the characters determining the high yield of the parents in these high yielding selections.