When crossbred offspring are used for production of an agricultural commodity profit is a function of three genotypes, that of the sire, the dam and the crossbred offspring. Heterosis in the component traits of profit, in any of the above genotypes, affects profit. Thus only when the final product is a four-way cross can we get crossbred parents plus a crossbred progeny and a full use of the available heterosis.

When the contributions of the sire and dam to profit are unequal, there is a justification for the breeding and use of specialised sire and dam lines, even though the component traits may be genetically additive. Multiple crosses (three- and four-way) may improve profit over two-way crosses only when there is heterosis in one or more of the component traits.

When profitability is presented as a function of productivity and repro-ductivity, five heterosis factors of the component traits may affect profit. These are: (a) Improved efficiency of production in a crossbred ‘commercial offspring’, (b) Improved viability (a component of reproductivity) in a crossbred ‘commercial offspring’, (c) Improved ‘female fertility’ in a crossbred dam. (d) Improved efficiency of the ‘commercial offspring’ through heterosis for maternal effects in the crossbred dam. (e) Improved male fertility through heterosis in the crossbred sire.

When the profit equation is non-linear, the same degree of heterosis as that measured on the component traits may have quite a different effect on profit, depending on the level of performance of the traits. By crosses and backcrosses between the parental lines the location and composition of the heterosis factors may be altered favourably.

Graphic and algebraic procedures were outlined for finding sire-dam combinations that maximise profit in the presence of heterosis when a group of genetic stocks is available.

Selection on an index of two traits was represented graphically by means of a response ellipse or response circle. This procedure was used to find optimal index weights for uncorrelated traits, and to compare the efficiency of alternative indices.

Profit was expressed as a reciprocal function of reproductive performance, from which it was shown that the higher the reproductive performance, the smaller the weight that should be applied to it in a selection index. It was found that in pigs and broilers the average commercial standard of reproductive performance is sufficiently high that selection on production traits alone is almost as efficient as selecting on the optimum index.

Three alternative procedures for selecting from a single original line were compared graphically: A. Maintaining the single line and selecting all animals on the same index. B. Maintaining the single line but selecting males and females on separate indices. C. Splitting the original line into separate sire and dam lines, each selected on a specialised index. Method C was found to be the most efficient of the three; B was more efficient than A for one generation, but less efficient in subsequent generations.

1. Twelve gilts, six pregnant and six non-pregnant, were each given a daily ration of 2.2 kg. for a 110-day period.

2. Live-weight gain during this period was 53 kg. for the pregnant animals compared with 23 kg. for the non-pregnant animals.

3. Apparent digestibility of nitrogen and dry matter and the energy of digested feed did not differ significantly between the pregnant and non-pregnant animals.

4. Daily nitrogen retention was on average 9·2% higher for the pregnant animals, the difference between the two groups becoming progressively greater during the trial.

5. Highly significant differences were found in pre-slaughter live-weight, weight of reproductive tract and weight of mammary region. The difference in the weight of free draining blood was also significant. No difference was found in the weight of alimentary tract, visceral organs, head or decapitated carcass.

6. The carcasses of the pregnant animals contained less subcutaneous fat plus skin and more intermuscular fat plus muscle than those of the non-pregnant animals. The weights of bone did not differ.

An analysis has been made of the reproductive performance of a breeding flock of Clun Forest ewes in Northumberland. Data are based upon 941 breeding females and 75 males. The average lambing percentages were 43·9 115·4 and 137·7 for ewes aged 1, 2 and over 2 years. Total lamb mortality to 3 or 6 weeks of age was 23·7% for lambs from 1-year-old dams, 16·6% from 2-year-olds, and 12·9% from older ewes. Mortality for all ages of dams was 12·1 % for single lambs, 15·0% for twins and 34·0% for triplets. Birth weight was strongly associated with survival within all birth types and maternal ages.

Analyses of weights at birth, 21, 42 and 112 days showed that much of the observed variation in lamb weight was associated with age of dam and type of birth and rearing. Single lambs and those from mature dams were significantly the heaviest. The effect of age of dam on lamb weight decreased sharply from birth to weaning, whereas the type of birth and rearing was the major influence at 21 and 42 days, and was still important at 112 days.

The mean weights of all lambs adjusted by least squares analysis to the basis of male twins from 3-year-old dams were 9·2, 19·7, 30·6 and 62·7 Ib. at the four ages. Four- and 5-year-old ewes reared the heaviest lambs. Interactions between the effects studied were significant both within and between years and accounted for from 7 to 10% of the total variation. A single set of general correction factors at each age was shown, on average, to account for 80–90 % of the variation between subclasses within years and may have applicability to comparable flocks.

An experiment was conducted on 13 commercial barley beef units in Scotland, using a total of 301 early-weaned Friesian male calves. On each unit the trial batch of calves was divided into two groups, which were then allocated to a control diet (85% barley + 15% protein supplement) or the same diet supplemented with 20 mg. chlortetracycline per kg. of feed. The experiment was conducted over the live-weight range 115 to 360 kg.

The chlortetracycline significantly (P<0·01) improved live-weight gain ( + 4·9%) and feed utilisation ( + 4·8%). This overall effect was largely due to a high response during the first 12 weeks. From 13 to 28 weeks the chlortetracycline had only a marginal effect, which was not significant. There was no significant difference in the feed intakes of the two groups. There was a significant (P<0·01) negative correlation between the effect of the chlortetracycline on live-weight gain at each unit and the growth rate of the control animals at that unit; the response was greatest on farms with poor hygiene. The animals given chlortetracycline showed a significantly (P<0·05) lower incidence of liver abscesses (11·8%) than the controls (28·2%). Macroscopic examination of the rumens showed that rumenitis is common among intensively fed cattle, but no correlation could be found between incidence of liver abscesses and rumen lesions.

Thirteen Blackface and Merino × Cheviot female sheep were shorn and each exposed for a maximum of 10 hours to intense cold (—18° C, 4 m.p.h. wind). Individual exposures were terminated earlier if rectal temperatures fell by 4·5° C. In the first year of the experiment (1963) 11 out of 13 of the sheep were in late pregnancy. There were also 12 control sheep, 9 of them pregnant, which were not subjected to cold treatment. In 1964, 12 of the previously treated sheep were subjected again to similar cold exposures. This time they were unmated and there were no controls.

The cold treatment apparently caused increased variation in gestation length, but the effects on neo-natal lamb mortality were inconclusive.

In 1963 and 1964 there was wide variation between individual sheep in cold tolerance—defined as the ability to maintain rectal temperature during cold exposure. Repeatability for this character was approximately 0·73, suggesting the existence of a genetic component. The average decline in rectal temperature was similar in both years: 3·9·2° C. in 1964. Pregnancy and breed had no detectable effects on cold tolerance.

Skin temperatures on the midside, ear and hind foot (pastern) were measured in 1964. They all declined during cold exposure. Midside skin temperatures typically fell to about 10° C., ear temperatures to about 3° C. and foot temperatures to zero. There were temperature fluctuations due to periodic vasodilation on the ears but the feet generally showed sustained vasoconstriction.

Superior cold tolerance tended to be associated with low mean body skin temperatures and a previous history of live-weight gain. It is suggested that subcutaneous fat deposition may be one factor responsible for improving skin insulation and thereby reducing heat loss in shorn sheep, but direct evidence for this was lacking.

Curves of growth in live-weight, from birth to 16 weeks of age, of 362 Suffolk × Welsh lambs were analysed for the effect of sex, birth type, sire and test centre environment and related to variation in age at slaughter.

The relative pre-natal disadvantages of twin lambs resulted in their curves of growth differing from those of single-born animals, particularly in males. Test centre environment exerted a marked effect in such cases. At one centre, male twins showed marked compensatory growth.

Differences in growth existed between the 14 sire progeny groups. As the linear component of the growth curves decreased, the quadratic values became increasingly negative. Deceleration of growth was most marked in slowest growing progeny groups and this was accentuated by a poor test centre environment.

These factors combined to influence age at slaughter, which was also related to birth weight of the lamb. Lambs at the best test centre reached slaughter weight 10 and 16 days earlier than lambs at the two remaining centres. Sires produced differences in age at slaughter of 10 to 14 days, within centres.

In a study of 100 pig carcasses, representing a sample of the pigs used in previously reported work, gross chemical composition of the ham, shoulder, loin, belly and half carcass were determined for one side. Percentage yield (trimmed v. untrimmed) of picnic, butt, ham and loin (PBHL), and certain carcass measurements were recorded on the other side of the carcass. From the covariance analyses, it was concluded that backfat thickness of the carcass alone was of only moderate predictive value for total chemical carcass fat.

Carcass protein could, for most purposes, be predicted with satisfactory precision from the percentages of protein in the loin or in the ham, with residual standard deviations of 0·27 and 0·36% respectively.

Although percentage yield of lean cuts may be considered satisfactory for commercial evaluation of swine carcasses, it was not an accurate index of the true leanness of a carcass expressed as the percentage of chemically determined protein.

Multiple regression equations were computed for predicting the percentage of protein in the carcass and percentage yield of PBHL from conventional carcass measurements. The residual standard deviations for predicting the percentage of protein were 0·49 for the males and 0·42 for the females; for percentage yield of PBHL these were 1·31 and 1·22 respectively. Total backfat thickness was the most important single variable in predictions. Variances among sexes were non-homogeneous for the important traits, with the females in all cases showing the greater variability. The results of regression analysis within sex indicated that, for maximum precision, a separate prediction equation for each sex was required.

1. Thirty-one gilts were injected subcutaneously with 1,500 i.u. PMS on Day 15 of the cycle, and inseminated on the morning of the 2nd day of oestrus.

2. Twenty of the animals were killed 1 to 4 days after insemination to study fertilisation, and the remaining eleven were killed on the 25th day of pregnancy to study embryonic survival.

3. All animals exhibited normal oestrous behaviour, only two of them failing to return to oestrus by Day 19 of the cycle. Treatment with PMS shortened the follicular phase of the cycle in nearly all instances.

4. Eighteen of the twenty animals in which eggs were studied were fertile and within these 84·2% of the eggs examined were undergoing normal development, 2·1% were fragmenting, 4·0% were polyspermic and 7·8% were primary oocytes. Ten unpenetrated secondary oocytes (1·9%) were also recovered.

5. Of the eleven animals in which embryonic survival was studied, only eight were pregnant at autopsy; the remaining three had 35 to 44 corpora albicantia, but no embryonic remains could be detected in the uterus.

6. The average ovulation rate in the eight pregnant animals was 35·4, with a range of 13 to 48. However, the average number of normal embryos was 17·0, with a range of 7 to 28.

7. The percentage survival of normal embryos varied from 26% to 83%. Embryonic survival was better at ovulation rates below that of the mean (35·4); an abnormally large or complete loss of embryos occurred at ovulation rates higher than 36.

8. Despite the range in percentage survival, the number of normal embryos found at the 25th day of pregnancy in five of the gilts was considerably above that found in untreated gilts.

1. The relationship between first lactation performance and the ability of individual animals to survive to the second lactation has been examined in Friesian and Ayrshire populations in an attempt to isolate the main determinants of survival, and to rationalise an apparently complicated selection procedure.

2. Low milk yield was the most important single cause of culling in the first lactation, though the two breeds differed considerably in the relative importance of milk yield in the selection programme. Natural wastage, caused by factors beyond the control of the farmer, accounted for the disposal of 5·7% of the population during the first lactation. Selection for butterfat was marginal in both populations, and could only be detected at extreme butterfat percentages, where high butterfat tended to compensate for a poor milk yield, and low butterfat to confirm a decision to cull an average yielding animal. Survival was not significantly affected either by the size of the herd or by the age at calving.

3. The proportion culled, and the intensity and effectiveness of selection for milk yield, were inversely related to the level of herd performance, and were greater in summer calving animals than in winter calvers.

4. The analysis indicated that selection for milk yield in different environments could be adequately described in terms of the Normal Curve. Yield-survival curves were computed and compared for groups of animals in different herd and calving season environments. When milk yield was expressed in terms of the deviation from the herd-year average, it was found that summer calvers, and animals in low yielding herds, had to demonstrate a productive capacity of far greater individual merit, if their chances of survival were to match those of their winter calving herdmates, and those of their contemporaries in higher yielding herds. When milk yield was expressed in absolute terms the survival of individual animals of similar yield was comparable in all herd and season environments. It was concluded that all the herds included in the analysis had been subject to the same absolute standards of milk yield in the selection of first calvers.

Selection procedure has been examined in Friesian and Ayrshire populations of dairy cattle in an attempt to determine the extent to which the relatively simple selection process in the first lactation was modified by the effects of age, by the incomplete repeatability of milk yield, and by the availability of production information from more than one lactation.

As a result of a greater intensity of selection for milk yield in the lower yielding herds, effective selection for yield was maintained at the later cullings despite increased wastage for reasons other than yield in the ageing population.

The high yielding animals that were selected by farmers in early life continued to outlive and to outyield their lower yielding first lactation contemporaries, notwithstanding a small amount of evidence that the heaviest yielding group were severely culled in the second lactation, mostly for reasons beyond the farmers' control. There was also some evidence that a high first lactation yield predisposed the animal to disposal for reasons other than yield in later life.

Survival at any particular culling was primarily determined by the milk yield in the previous lactation. In this respect, the culling after each lactation appeared to be independent of previous selection decisions, and the individual animal had to demonstrate its merit repeatedly to survive—earlier merit did not compensate for a subsequent failure.

The investigation was designed to measure the time taken to eat concentrate foods and to determine whether the type and quantity of food, the age of the cow, the length of time between feeds and the eating habits of the cow influence consumption rate.

The data indicate that the main factor affecting eating time is the quantity eaten. Within the range of 2 to 8 Ib., the larger the ration the faster the rate of eating. Further, the data show that up to 20 Ib. of concentrates can be consumed per cow per day in a parlour without any check in the faster rate of eating recorded when the higher levels of concentrates are offered.

The type of concentrate is also important: cubes are eaten faster than meal. There are considerable variations between cows in the time taken to eat.

The results of this investigation did not show any obvious effects on the time of eating of breed of cows, position of manger, nor of the length of interval between bulky and concentrate rations.

Graphs relating the eating times of cows with the times available for eating in parlours are shown, and these can be used for reference on farms in parlour planning.

1. Twelve groups of 3 litter-sister Large White gilts were given during three successive pregnancies a daily meal allowance of either 6 lb. (Group A) 3 lb. (Group B) or 3 lb. for 76 days followed by 6 lb. until parturition (Group C). During 8-week lactation periods all received the same daily allowance of 4 lb. meal plus 0·8 lb. for each piglet suckled.

2. Net weight gains during the three successive pregnancies respectively were: Group A, 114, 97 and 79 lb.; Group B, 11, 10 and 7 lb.; Group C, 47, 44 and 35 lb. The differences between groups were highly significant (P<0·001) and there was a significant parity interaction (P<0·05) in that Group A sows showed a significant decline in net weight gain from parity to parity whereas those of Groups B and C did not. No relationship was found between net weight gain and litter weight at birth.

3. Weight changes during the three successive lactation periods respectively were: Group A, –37, –41 and –36 lb.; Group B, 2, 14 and 4 lb.; Group C, –10, –14 and –5 lb. The differences between groups were significant to various degrees in different lactations. There was no relationship between weight changes during lactation and total litter weight at 3 weeks of age.

4. Overall weight changes from first service to third weaning were: Group A, 160 lb.; Group B, 72 lb.; Group C, 71 lb. The differences between Group A and Groups B and C were highly significant (P<0·001).

5. The implication of the results as an indication of feed requirements for reproduction is discussed.

The importance of fat thickness and size and shape of eye muscle in bacon pigs has initiated much work to establish convenient routine measurements that would most reliably estimate these qualities. Measurements taken of the muscle: fat ratios in the region of the loin have been found to be the best estimate of the whole carcass and (as discussed by Harrington, 1958) have been a valuable aid in carcass competitions, where, to ensure similarity between contests, judges have used standard photographs of cut sides. The use of routine photographic records of cut sides in nutrition and breeding work has not, however, been fully explored, but was used by American and German workers (Ferguson, Brennan and Buric, 1961; and Brohmann, 1961). The advantage of using photographic records is that the measurements can be made at the most convenient time and place and form a near permanent record. They can also be used in developing correlations with measurements on live pigs which might estimate the quality of the rashers later produced.

During the last 10 years many genetical polymorphisms have been discovered in blood proteins. The phenotypes within each system have usually been found to be determined by a series of two of more alleles at a single locus and between which there is no dominance. A number of investigations have compared the fitness of heterozygotes and homozygotes at these loci. Apart from examining for possible associations between heterozygosity and traits for which selection is likely, such as production characters in domestic animals, these investigations have taken one of two forms; either family data have been examined for agreement with Mendelian ratios or else the genotypic numbers in a population have been examined for agreement with Hardy-Weinberg expectations.