When planning a factorial experiment, it is often desirable (as pointed out by Yates (1937)) to include certain extra treatments falling outside the usual factorial scheme. I discuss here the exact analysis of an experiment of this type. Only standard principles of experimental analysis are involved, and the full exact analysis is not always required in practice, but the example may illustrate techniques that are useful in investigating more complicated designs that are sometimes encountered. I am grateful to the late Dr E. M. Crowther and Miss B. Benzian, of the Chemistry Department at Rothamsted, for discussion of the problem and for providing data for the illustrative example.

1. An investigation to compare winter and spring defoliation of S.147 oats, and incorporating an application of nitrogen, was carried out during the seasons 1950/1, 1951/2 and 1952/3.

2. Forage available in the spring was considerably greater than that available in the winter, but it had a lower crude-protein content.

3. Compared with the undefoliated oats, both spring and winter defoliation reduced the number of productive tillers by significant amounts, but nitrogen had no effect on tiller numbers under defoliated and undefoliated conditions.

4. Spring and winter defoliation caused significant reductions in grain yields; with winter defoliation the decline in yields was substantial. On the undefoliated and spring-defoliated oats nitrogen increased yields by significant amounts.

5. Spring and winter defoliation reduced straw yields, but nitrogen on both treatments and on the undefoliated oats gave significant yield increases.

1. An investigation is described in which the effect of both technical and physiological factors on muscle fibre size was examined. Size was determined by measuring the cross-diameter of 16,450 individual fibres by means of an ocular micrometer. In cases where an animal was represented by a single muscle sample, 100 fibres were measured, otherwise fifty measurements were recorded per sample.

2. The material consisted of muscle samples always taken from the same position along the length of the muscle and immediately afterwards fixed for varying periods of time in 10% formalin. Samples were obtained from various sources, principally from experiments conducted in the past at the School of Agriculture, Cambridge, but also from slaughterhouses and contemporary investigations. These latter samples were treated in the same manner to ensure consistency.

3. From the results of an investigation on the effect of formalin fixation on muscle fibre diameter, carried out on samples obtained from a mature buck rabbit, it was tentatively concluded that although shrinkage does occur the effect is only slight; the difference between measurements of fresh compared with fixed fibres being non-significant statistically. Shrinkage apparently ceased as soon as the fixative had penetrated the sample, since no further changes could be detected after prolonged fixation.

4. The same material was used to study the effect of heat. Application of heat (boiling) for a matter of seconds caused an immediate, highly significant, shrinkage.of individual fibres, but continued heattreatment resulted in no further changes. Though heat caused the fibre to lose its characteristic striated appearance, there were no signs of fibres disintegrating or of the sarcolemma becoming detached from its protoplasmic contents.

5. From the results of previous investigations it was deduced that the measurement of 100 fibres per sample provides a reliable mean value and a representative indication of the dispersion in a given muscle, or at a given age. Results obtained in the present study demonstrated a slight tendency for larger fibres to be measured as the period of observation proceeds, a factor which should be guarded against.

6. The effect of species on muscle fibre diameter was examined by comparing fibres of m. gastrocnemius of the rabbit, the pig, the sheep and cattle at birth and maturity. Only male animals were included in the comparison. It was shown that no relation exists between muscle fibre size and body size at either age. At birth the rabbit and sheep had rather similar-sized fibres, while those of the pig and cattle were respectively smaller and larger in diameter. At maturity the pig had the largest fibres, followed in decreasing order by the rabbit, cattle and sheep. The size of muscle fibres at maturity was accounted for by the degree of post-natal development in body weight of the particular species.

7. The effect of breed was studied on two groups of steers, the one consisting of Dairy Shorthorns and Dairy Shorthorn-crosses, and the other of Friesians and Friesian-crosses. Samples were available from m. longissimus dorsi for each of thirty-four animals: 100 fibres were measured per sample. The Friesians and their crosses had significantly thicker muscle fibres than the pure- and cross-bred Dairy Shorthorn steers. The largest differences existed between the respective pure-bred animals; differences between Friesian × Angus and Dairy Shorthorn × Angus, and Friesian × Hereford and Dairy Shorthorn × Hereford, though fairly distinct, were, however, not statistically significant.

8. The effect of age was investigated on a group of forty-one lambs of different nutritional status and sex, and ranging in age from birth to 290 days. Muscle fibre diameter was shown to increase in general with age, while a consistent decline in the coefficient of variation was regarded as indicative of the fact that muscular growth during post-natal life occurs essentially by hypertrophy of individual fibres, there being no increase in the number of fibres after birth.

9. Correlating changes in muscle fibre diameter with corresponding changes in weight, indicated that muscular growth is primarily a function of physiological age, and not strictly one of chronological age. Though highly significant correlations were established between mean fibre diameter and body and carcass weights, the strongest correlation was shown to exist between the former variate and muscle weight. However, a correlation of an even higher order was obtained between the square of muscle fibre diameter and muscle weight. It was attempted, by means of linear regressions, to indicate the contribution of length growth of the fibres to increments in muscle weight. The need for further investigation is, however, apparent; insufficient data in the present study made it impossible to elucidate this point altogether.

The relationships were confirmed by an analysis of twenty lambs of the same breed, all slaughtered at 112 days of age; the heavier lambs had larger fibres than their lighter counterparts, very nearly proportional to differences in weight of muscle.

10. Sex differences in muscle fibre diameter could very nearly be accounted for on a basis of muscle weight alone at birth and at a carcass weight of 13·6 kg. At 290 days of age, high-plane wethers had but slightly thicker fibres than their female counterparts, despite a significantly heavier musculature. This was ascribed to differences in length of muscle (as shown by bone measurements), and also to differences in composition of the muscle. The results of chemical analyses were presented to prove that the muscles of wethers at that age contain greater amounts of intra-muscular (chemical) fat, hence the apparent increases in the weight of muscle could not be accounted for by an increase in the diameter of component fibres.

11. The effect of nutrition was studied in both lambs and mature ewes and shown to influence muscle fibre diameter appreciably at all ages. However, at birth the differences, though in favour of the high-plane lambs, were not significant statistically, probably due to small numbers in the respective groups. Muscle fibre diameter of mature ewes on a supermaintenance diet increased in proportion to increases in total muscle, while on a submaintenance diet the opposite effect was found. It appeared that continuation of the supermaintenance treatment would have resulted in but little additional changes in the diameter of fibres, while prolongation of the submaintenance treatment probably would have caused considerable further shrinkage of individual fibres.

12. The effect of the individual muscle was studied by comparing absolute and relative development of fibres of m. longissimus dorsi, m. rectus femoris and m. gastrocnemius. At birth m. gastrocnemius possessed the largest fibres and m. longissimus dorsi the smallest. On the whole, fibres of m. longissimus dorsi, the muscle being situated in a late maturing part of the body, showed greatest relative increases during post-natal life, while those of m. gastrocnemius, an earlier maturing muscle, increased latest. On an average, m. rectus femoris had larger fibres than m. longissimus dorsi at maturity, those of m. gastrocnemius being the smallest in absolute measure. Comparing the relative degree to which different muscles develop under high and low planes of nutrition, muscle fibre measurements indicated that the low-plane animals at a chronological age of 290 days, resembled their 60-day-old high-plane counterparts in anatomical development.

In mature animals, the early maturing gastrocnemius appeared to benefit most initially from a supermaintenance diet; however, during the later stages of the experiment, m. longissimus dorsi fibres showed the greatest relative increases, followed in order by m. rectus femoris and m. gastrocnemius. On the submaintenance ration, m. longissimus dorsi fibres appeared to be reduced in size at a greater rate initially than those of the other muscles. Individual variation, however, made it extremely difficult to generalize during subsequent stages.

13. From width and depth measurements on m. longissimus dorsi (or ‘eye muscle’) recorded at the junction of the thoracic and lumbar vertebrae, it was shown that width is the earlier maturing dimension and less affected by nutritional factors than depth.

Though significant relationships were found between mean fibre diameter and both muscle width and depth, the latter dimension was the more strongly correlated with changes in thickness of fibres.

14. It has been suggested that work of this nature might provide a suitable basis for estimating the amount of muscular tissue in a carcass; such a relationship would be of great assistance to students of meat physiology who have to resort to laborious dissection techniques for data on carcass composition. It is obvious, however, that factors such as species, breed, and possibly sex, would have to be considered in an attempt to establish a relationship of this kind. Furthermore, the effect of the individual muscle used for test would demand consideration, different muscles being influenced to varying degrees by age and nutrition. A late maturing muscle probably would furnish the most reliable criterion; in the light of evidence produced by this study, m. longissimus dorsi sampled at the junction of the lumbar and thoracic regions has been suggested as most suitable.

1. A series of experiments, involving the use of three hundred and forty Rhode Island Red x Light Sussex day-old cockerels, was carried out to determine (a) the influence of floor type on the incidence of perosis, and (b) a method of lowering the high incidence of perosis which may occur when chicks are fed the experimental diet no. 124 originally described by Dam et al. (1951).

2. It was found that the incidence of perosis was about 60% higher in groups of chicks raised on wiremesh floors than in groups of chicks similarly raised on wooden floors.

3. Neither the area of the floor space nor the level of manganese in the diet influenced the incidence of perosis.

4. When the calculated choline content of the diet was 500 mg./kg. the incidence of perosis was 60% but when it was raised to above 800 mg./kg. the incidence fell to about 15%. This could not be reduced further by further increasing the choline content.

5. Since the rate of body-weight gain increased as the choline content of the diet was raised to 1400 mg./kg. it is suggested that diet no. 124 may be improved from both the points of view of perosis prevention and of growth rate if the choline content is raised to 1500 mg./kg. of diet.

1. Eighteen determinations of the energy retention of six sheep were made when they were given the same batch of dried grass in the form of chopped material or as cubes. The cubes were made following hammer milling to a medium and fine particle size. The fasting heat production of each sheep was also determined, following subsistence on a standard ration.

2. Agreement between determinations of energy retention calculated from the carbon and nitrogen retentions and from the energy exchange was good. The mean discrepancy was 4 Cal./24 hr.

3. There were no statistically significant differences in energy retention as between the three materials at either a low (600 g./24 hr.) or a high (1500 g./24 hr.) level of feeding. Calculations of net energy/100 Cal. of food ingested showed that higher values occurred at the lower level of feeding. Standard errors of the means were small, about ±3% of the determined values. Further analysis showed that no large differences in the net energy value of the materials would appear within the normal feeding range, but slight extrapolation of the data indicated that the cubes would be superior at high feeding levels.

4. Faecal losses of energy were considerably greater when cubes were given and methane losses were much smaller. Individual sheep which showed low methane losses also showed high faecal energy losses. Faecal losses of energy were smaller at the lower feeding level. Urine energy losses were unaffected by the amount or physical form of the food given.

5. Heat losses were greater at the higher nutritional level and were considerably less for cubes than for chopped material. Constancy of net energy value in this study thus involved compensation of high faecal energy losses by low losses of energy as heat and methane.

6. The determinations of the digestibility of the carbohydrate fractions of the grass showed that a fall in the digestibility of the structural components of the cell was the major factor causing increased faecal losses. The digestibility of intracellular constituents fell very much less.

7. It is shown that evaluation of the grasses in terms of metabolizable or digested energy does not place them in their correct physiological order of nutritive value, and that estimates of nutritive value using Kellner's and other factors do not give their true nutritive worth.

8. It is pointed out that physical factors, which change the rate of passage of food through the gut, change the rate and nature of the microbial fermentation, and cause variation in the mechanical work involved in prehending, masticating and cudding food, are as important as the chemical composition of the food in determining its nutritive value.

The fleeces of Border Leicester, Romney Marsh, Lincoln and Dorset Horn ewes which were confined for 3 years to copper-deficient pastures at Robe, in South Australia, have been studied.

As the copper status of the untreated ewes fell, their fleeces revealed abnormalities identical with those encountered in similarly copper-deficient Merino and Merino crossbred sheep.

The crimp characteristic of each breed was lost and straight, ‘steely’ fibres were produced. The capacity for crimp formation returned immediately when the copper status was restored.

Ewes which received regular copper supplements grew normal wool throughout the experiment. Supporting analytical data are presented.

The absence of this specific lesion from the fleeces of copper-deficient British breeds of sheep elsewhere is discussed.

Using two inbred lines of Brown Leghorns, B and I, a backcrossing experiment was set up to test whether inherent rate of production was a simply inherited trait. Female progeny of each generation, selected for high rate, were mated to males of the B (lower rate) line. The pullet records now available, for 2 full years' production and to 31 March in the third generation, have yielded unexpected results.

In each backcross generation mean egg production has been at least as good as in the better parent, line I. This is not wholly accounted for by the behaviour of rate, since measurements of the trait in the cross approximate to the average of the two inbreds in each year. High viability and absence of very poor layers were also determining factors.

Backcross egg weight remained intermediate between the widely differing parent lines, and the cross produced the greatest total weight of egg material in each set of comparisons. Body weight was also intermediate, but sexual maturity was comparable with that of line B.

It was deduced that selecting the backcross breeding females for high rate had led to the retention of a greater proportion of the original I genotype than would normally have been expected, and so maintained continued heterozygosity in succeeding generations.

Practical implications of the results and interpretation are discussed.

Thirteen experiments on spring cereals were carried out from 1952 to 1954 to compare broadcasting with combine-drilling for nitrophosphate and for mixtures of superphosphate plus equivalent nitrogen. Broadcast nitrophosphate was of little value; drilled dressings were much more effective. Superphosphate drilled with the seed produced consistently higher yields than broadcast superphosphate. Superphosphate plus equivalent nitrogen gave higher average yields than nitrophosphate both when broadcast and when drilled. When the dressings were broadcast, superphosphate was markedly superior to nitrophosphate, but when the fertilizers were drilled with the seed, average yields with superphosphate were only a little higher than with nitrophosphate.

Nitrophosphates may be substituted for mixtures containing water-soluble phosphorus for cereals providing that they are drilled with the seed. Since nitrophosphates do not make cereals grow as rapidly as does superphosphate, even when drilled with the seed, such materials may give poor crops in conditions where the plants need a rapid start. The nitrophosphate tested contained rather more than one-quarter of its phosphorus in water-soluble form; the improved early growth observed where nitrophosphate was combine-drilled may have been largely due to this fraction of the fertilizer.

1. The results are presented of two series of digestibility trials. The first series consists of six grass hays, cut at different stages of growth, and two legume hays cut at the flowering stage. The second series consists of trials using seven species of grasses at the ‘standing hay’ stage.

2. Good-quality grass hays are obtained from grasses at the flowering stage. More mature grasses yield poor- or very poor-quality hays of low-protein content.

3. The legume hays were of good quality, being particularly high in digestible protein.

4. The grasses at the ‘standing hay’ stage are generally of poor feeding value, Brachiaria brizantha, a drought-resistant species, being the best of the seven studied.

5. From these results a tentative conclusion is drawn that, during the dry season, the extensive grazing areas of East Africa are capable of providing, at the most, for the maintenance requirements of ruminants.

1. An examination has been made of the results of a number of experiments on the feeding of fodder beet to fattening pigs.

2. When beet is introduced, the daily energy intake and hence the growth rate are slightly reduced.

3. The value of beet as a replacement for meal, neglecting the reduction in growth rate, may be assessed satisfactorily from a knowledge of the dry matter content of the beet and of the starch equivalent of the meal, using the average values given by Woodman (1954).

4. Where beet is used to replace meal, the available digestible protein is reduced. It is possible that the addition of protein-rich supplements to the ration will improve the performance of the pigs.

5. The effects of beet on carcass quality are not clear. The Danish reports examined indicate slightly lower quality, but the English reports similar or slightly higher quality.

The ability of various high polymers to stabilize soil crumbs has been investigated using the sodium saturation technique. The soil used came from two plots on the Rothamsted classical fields, Barnfield and Hoosfield, which have received no fertilizer either organic or inorganic for 80 years, but whereas the former contains free CaCO3 the latter is acid.

The polymers tested were poly vinyl alcohol, sodium alginate, a vinyl acetate-maleic acid co-polymer, polyacrylic acid and a dextran. It was necessary to add 0·2–0·5% weight of the polymers in solution to the calcareous soil to produce crumbs of maximum stability, i.e. not dispersing the distilled water. The alginate, however, was effective only if the crumbs were first sodium saturated. With the acid soil, the amount of polyvinyl alcohol necessary for maximum stability was slightly reduced, corresponding to the lower clay content of the soil, whereas only one-tenth the concentration of the maleic acid co-polymer was required. Sodium alginate had no effect on the acid soil.

From the swelling pattern of the sodium-saturated crumbs in dilute salt solutions it was inferred that the non-ionic polymers, polyvinyl alcohol and the dextran form interlamellar complexes, whereas the carboxylated polymers are joined to the edge faces of the crystals. The presence of divalent cations is unnecessary for this edge linkage, as the polymers are equally effective when sodium ions only are present. It is suggested that the polymers form a series of hydrogen bonds with the exposed oxygen and hydroxyl atoms of the octahedral layer. Since these atoms have a greater tendency to co-ordinate a hydrogen ion as the pH is lowered, the carboxylated polymers would be expected to be more efficient on acid soils.

A simplification of the sodium saturation technique for the evaluation of potential soil conditioners is given.

Evidence obtained in the Agricultural Improvement Council Survey of Maincrop Potatoes 1948–50 and in experiments shows that yields in England and Wales are decreased by delay in planting after about 11 April at the rate of about 0·4 tons per acre per week. The effect is rather greater in years with fine springs, and on high-yielding fields.

In Rothamsted experiments the responses to dung and fertilizers were all greatly reduced when planting was delayed; between early April and late May the effects of dung, nitrogen and phosphate were halved, while that of potash was reduced by about 80%. The evidence on the responses of different varieties to early planting in English experiments is contradictory, except for an indication that higher yielding varieties respond better. Data from Craibstone suggest that, if bulk yield is the only criterion, ‘early’ varieties which make rapid growth early in the season have a later optimum planting date than maincrop varieties.

The effect of chitting seed at Craibstone is to increase yields by 1·7 tons per acre for March plantings, and by 2·5 tons per acre for plantings in mid-April and later. There is negligible loss of yield from sprouted seed at Craibstone for plantings delayed up to mid-April, but for unsprouted seed there is evidence of a gain from plantings a month earlier. Sprouted seed planted even as late as 1 May yielded as well as unsprouted planted two months earlier.

1. 863 female pigs, mostly sows discarded from commercial herds in East Anglia, and sold for slaughter, were examined when killed and the condition of the reproductive tract was related to the animal's reproductive history when it was known.

2. Anatomical abnormalities of the reproductive organs, other than cystic ovaries, were found in sixteen animals (including two gilts) and the abnormality involved the ovary in eight of them. At least two of them were brought to notice because of the abnormalities so that the number seen affords a maximal estimate of the incidence of this type of aberration in the pig population from which the sample was drawn.

3. A wide variety of ovarian cysts was found, often within a single pair of ovaries. It is suggested that all the ‘types’ of cyst described here and elsewhere are different degrees of the same kind of aberration, the morphological form being determined by the stage in the ovulation process reached by the follicle (or corpus luteum) when overtaken by the physiological breakdown responsible for the cystic distension.

4. As none but slight degrees of cystic abnormality were found in pregnant animals, these were used as a criterion to distinguish between slight and severe degrees of cystic abnormality, the latter in all probability associated with sterility, and encountered in about 10% of the sows in the sample.

5. Slight degrees of cystic abnormality do not appear to be associated with the production of small litters; pregnancy appears to be either unaffected or else precluded altogether.

6. The reproductive organs of more than half of the sows which were discarded for reproductive failure were found to be normal, and in many cases the failure could only be attributed to chance. A large proportion of the sows discarded as sterile were culled after their first litter. The preponderance of reproductive failure at this stage could not be attributed to ovarian cysts.

7. There was a marked seasonal variation in the incidence of ovarian cysts, the proportion of animals with some degree of cystic abnormality being twice as high in the spring months as in the autumn. The variation is thought not to be due to sampling errors alone. There was no seasonal variation in the average number of corpora lutea or of embryos.

8. The observations are discussed in relation to work of a comparable nature done elsewhere, particularly in U.S.A. Some striking divergences are apparent. Large cysts, commonly found in pregnant animals in America, did not appear to interfere with gestation and were evidently formed during pregnancy. Such cysts were never observed in pregnant animals in the English material, where only nine out of 130 pregnant animals showed any degree at all of cystic abnormality, by no means severe in any of them, and not involving ‘large’ cysts. Marked enlargement of the clitoris was found to be associated with one type of ovarian cyst in America but was not observed in England. A greater diversity of cystic abnormality was recognized in the present work than in American studies, and such histological and endocrinological work as has so far been done has given results somewhat different from those recorded in America.

9. There is strong evidence that oestrus may fail to occur in the sow, so that the animal cannot be served. The condition is probably reported more frequently than it occurs, however, since reliance is often placed on examination alone for the detection of heat, without using a boar.

10. There is some evidence that oestrus and ovulation may get out of step in some animals and it is possible that the cause is related to that of ovarian cysts. Oestrus and service may occur during pregnancy.

In field experiments on vegetable crops dressings of appropriate fertilizers were placed 2 in. to the side of the seed and 2–3 in. below the soil surface, and compared with the same amounts of fertilizer broadcast and worked into the surface soil. The work was carried out on ordinary arable land carrying rotations which included vegetables.

Placing fertilizer gave higher yields of cabbage, lettuce, beetroot, onions, broad beans, runner beans and maize, than broadcasting. Fertilizer, however applied, had no regular effect on the yield of french beans.

Both placed and broadcast fertilizer were tested at two rates of dressing. On average of all experiments on each crop, placing fertilizer at the low rate gave higher yields than broadcasting at the high rate. Broadcast fertilizer had little effect on yields of runner beans and broad beans, while placed dressings gave marked increases.

In two-thirds of all individual experiments placing gave higher yields than broadcasting, and in one-third of the experiments placement was significantly better than broadcasting. There were no instances of significantly higher yields from broadcasting as compared with placing.

Placing fertilizer made most of the crops grow more rapidly in the early stages than broadcasting, this improvement being often reflected in earlier maturity. For cabbage, lettuce and runner beans the relative gains from placing, as compared with broadcasting fertilizer, were much higher at the first than at the second harvest.

By drilling fertilizer beside the seed of vegetable crops grown on ordinary soils, it is possible to economize in the dressings needed. In addition, some crops having placed fertilizer may be ready for market earlier. These advantages may be very profitable where high-value crops are grown and will justify the purchase of special placement drills.

A small series of field experiments on established Brussels sprouts and autumn-planted cabbages compared mid-season top-dressings of nitrogen fertilizer broadcast by hand over the whole soil surface, with dressings placed at one point beneath the surface and near to the plants. The dressings used had practically no effect on yields of Brussels sprouts. Top-dressings broadcast in spring increased yields appreciably in two experiments on cabbage, and at both centres placing gave lower yields than broadcasting.