IN breeding for improved egg production, as in breeding for improved milk yield, measurement can be made only in one sex. The male is always unknown, and this unknown 50% of the inheritance in each generation on the average drags back the progeny of a good producer towards the mean of the breed. It was found in practice, as was shown by the classic experiments of Pearl and Surface, that starting with an ordinary farm flock of birds laying about 100 eggs per annum and breeding only from the best layers made initial improvement, but that a ceiling of about 160 eggs per annum average was soon reached, above which selection in one sex only made no appreciable improvement.

In the past few years a remarkable new interest in the breeding of dairy cattle has come to be shown in Great Britain and, as a result, renewed attention has been focussed on the subject down for discussion to-day. I think there are two main causes of this new interest. The first is the certainty—for as far ahead as we can see—that milk is likely to remain agricultural product No. 1. With this has come the conviction that to ensure efficient production at the source we have to introduce more certainty into our methods of breeding for milk. The poor level, by any standards of breeding efficiency, of a great number of our dairy herds has been one of the revelations of the war years; there is now factual evidence for what has been suspected for a long time.

In New Zealand, herd recording, since the introduction of the Herd Improvement Plan in 1938, has been governed in policy by the Herd Recording Council in Wellington. In 1938 the decision was taken to extend the scope of the milk recording movement so that milk records could be used for:—

• (1) improvement through breeding;

• (2) studies on farm management, including the prevention of disease in dairy herds; and

• (3) studies on problems of nutrition on dairy farms, em-bracing not only the feeding of dairy cattle and young stock, but also such matters as the influence of soil types, pasture species, etc.

There are six Herd Improvement Associations and a representative of each Association is elected annually to the Herd Recording Council with full voting power.

It is only possible in the time available to treat the background to problems of production along the broadest possible lines. The subject is one well worth detailed attention, but it is questionable if sufficient knowledge is extant to fill in the wide gaps in the story of the evolution of our highly productive modern breeds from their ancestral types—the jungle fowl.

The history of domestication of the fowl covers a very extended period of time, and in it three main lines of development may be traced. The first of these is the use of the male fowl as a fighting cock. From the earliest times up to the present it has been used for this purpose and cock-fighting has been the incentive to selective breeding. In this case the skill of the breeder has been put to the severest of tests—the elimination in battle of those individuals lacking the necessary qualities to survive.

Official Milk Recording has been operated in Scotland since 1904. The Association came into being in 1914. The objects of the Association as laid down by the Constitution and Rules are:

• (1) To foster and improve the milking qualities of Cows in Scotland, and

• (2) To institute on a public and reliable basis, the taking of daily or periodic tests of Cows for quantity and quality of milk.

To meet the terms of this Constitution a system has been evolved whereby a Recorder visits the farms of milk recording herd owners at intervals of 24-28 days, weighs the milk of all cows in milk in the herd, and tests for butterfat in respect of all cows which are to be included in the record. While the milk of all cows must be weighed, butterfat testing in respect of all cows is not essential.

I Feel at a distinct disadvantage compared with the speakers this morning because the foundations on which to discuss breeding methods, and progress, are more clear-cut in the case of poultry than in larger farm stock. Poultry are fast-breeding—you can get a lot of offspring quickly if you want them—they can be, and to a great extent are, kept in fairly well-controlled conditions of husbandry, feeding, and management; and their performance is simply and accurately recordable. All these points are important—in contrast, for larger animals we lack a very, great deal of the factual information needed for a reasonable stocktaking of our assets, difficulties, and liabilities.

However, from what I know of the workings of our Executive Committee's minds, I believe I am expected to be provocative and to leave loose ends for open discussion. I propose therefore to try to outline the scope for breeding methods in improving our livestock and, if possible, to pick out some of the general principles that govern their use and effectiveness. In any case you will recall that some of the questions have been discussed at earlier meetings of this Society, such as the rôle of cross breeding in sheep, specialisation in pigbreeding, and recording systems in cattle. I ask you to allow me to generalize, and to note that the examples I quote are only examples—I make no pretensions at all to give a comprehensive or a detailed picture.

During the last twelve months short tours were made in Denmark, Canada, and the United States and certain impressions were collected, which allied to actual statistics show a picture of the present development of artificial insemination in those countries. The statistics relate mainly to a period ending in 1944, and 1945 figures are not yet available. The absence of a uniform method of collecting and tabulating data in each country makes an accurate comparison rather difficult, but the general trend of progress can nevertheless be determined. It is stressed that it is not easy to confine the subject strictly to progress without at times trespassing on the field reserved for other speakers.

The scope of this subject is so wide that no attempt will be made to deal with it in detail, but I will treat it in a general way. drawing largely from my experience at the Reading Centre.

The technical aspect of artificial insemination has now reached a stage of development where good results are almost a certainty, but more than good results must be achieved if an Artificial Insemination Centre is to be run satisfactorily. It is of great importance that a service be provided at an economic figure, and this can only be done if the administration and organisation of the Centre are on a sound business footing.

I intend to confine my remarks to the organisation of the individual Centre, and in case any of the audience are unaware of it, the national development of artificial insemination is controlled by the Minister of Agriculture under the Artificial Insemination (Cattle) (England and Wales) Regulations, 1943, and he has for his guidance the Central Advisory Committee.

This paper is an attempt to throw a little light on what is probably the most neglected, and yet most important aspect of Artificial Insemination, namely, the management of the bulls used for this purpose.

Evidence is becoming apparent that a number of bulls used for A.I. have been discarded or compulsorily rested because they refused service. As the number of Centres increases so will the number of bulls which fall into this category, and the problem is likely to be the greatest that A.I. has to face in the future.

It is rapidly becoming obvious that this failure of sex drive is not, as generally thought, primarily nutritional, but is in a large degree due to the lack of knowledge of the psychological process involved in the sexual act, and the consequent ill-handling of the animals during service. The knowledge available on the mental processes which take place is inconclusive and complicated, and involves androgenic and oestrogenic factors and their relationship with these processes.

By control of semen I mean those tests which are applied to a semen sample to determine whether the sample is likely or not to be fertile when used for insemination, and to determine the extent to which it can be successfully diluted. For those who are not familiar with the methods of artificial insemination, I will try to give a very general account of the principles of semen control and preservation, hoping that more technical matters may be brought up in the discussion.