neoessity for making it. Yet, clearly, the problem of development is largely one of filling "the vacuum between determinant and character" (DARLINGTON 1951). Nowadays the chromosome theory can be presented in much greater detail and with utter confidence, but its two main features remain the same. However, while the role of the chromosomes in heredity and development has been appreciated for a long time, the manner in which they perform their genetic and epigenetic functions has become amenable to critical investigation only in recent years. There is, therefore, still an unmistakable tendency to think of chromosomes in terms of the discrete threads of cell division and, in keeping with this conception, the chromosome cycle is gen erally considered in relation to the microscopically visible changes in morphology which occur during the mechanically active phases of mitosis and meiosis. Chromosome phenotype, however, changes not only during division but throughout the cell cycle. The changes which occur during interphase are, of course, scarcely revealed in morphological modifications of the restless "resting" nucleus. Consequently they are less obvious and correspondingly less amenable to investigation. This accounts for the concentration on the countable karyotype, with its visible properties of pairing and pycnosity, and the measurable movements of separation and segregation.

material can serve both autocatalytic and heterocatalytic functions. Thus not only is it unique in itB capacity for self-replication but its base sequence determines the specificity of proteins. And enzyme proteins are immediately responsible for the peripheral metabolism which enables the organism to impose its own kind of order on the raw materials it absorbs. The course of development is determined not only by the nature of the genetic material but by its over-all amount and the relative frequency of the different functional units. Differential rates of epigenetic activity matter also. In theory, therefore, differential development within or even between individuals could be determined by the differential replication of the various genetic elements or by their differential activity. And further variation could arise by the differential transmission of these elements between cells. Indeed it would appear that all these possibilities are exploited by living systems. If like is to beget like, however, any genetic change which occurs during development must be undone, or else germinal units preserved from change must be set aside. As far as is known, genetic changes, even those involving only quantity or relative amounts, are reversible to only a very limited extent so that a change once done cannot be undone. Consequently genetic changes during the development of presumptive germ-lines are either non existant or minor and confined to a small class of un aggregated deter minants."

Introduction When the study of heredity and variation first came to be treated as a scientific subject-and this, one must remember, was only just over a hundred years ago-there was an unfortunate separation between the disciplines of cytology and experimental breeding. This separation was based partly on a lack of understanding and partly on a lack of the desire to understand. Even WILLIAM BATESON, the first apostle of mendelism in England, had a blind spot for cytology and for many years dogmatically refused to believe that MENDEL'S determinants were transmitted and distributed by the chromosomes. This separation between cytology and experimental breeding is one which persists, in a measure, even today, simply because there are two quite different, though complementary, techniques available for the study of heredity and variation. On the one hand, one can study directly the structure and behaviour of the actual vehicles which transmit the genetic determinants from one generation to the next. This is the method employed by those who study genetics through a microscope. The alternative method is that used by the experimental breeder who, in default of being able to watch the hereditary factors segregate from each other directly, is obliged to examine the constitution of the germ cells indirectly by sampling, and usually at random, the products of a controlled mating.