May I begin by recalling that at the Copenhagen Conference five years ago, Tjio and Levan had just published their almost apologetic announcement that they could only find 46 chromosomes in cultured human somatic cells, and Hamerton and I presented evidence that there were 23 bivalents in spermatocytes, not 24.

Since that time there has been a very marked expansion of work in human cytogenetics. The reason is not hard to seek: the findings have been of significance in so many fields of medicine ranging from endocrinology, through paediatrics, gynaecology and mental diseases to haematology and cancer.

To the biologist, this outpouring of interest may seem at times to have bordered on the hysterical. For him the developments may have done little more than illustrate in our own species the principles that were established in the twenties and thirties with other material. On second throughts he will realize that man offers unique advantages for cytogenetic study in virtue of the enormous numbers of individuals who are subjected to medical scrutiny and thereby become candidates for chromosomal examination. But he will also remember the serious disadvantages, (some shared with human Mendelian genetics) of long generation time, small family size, impossibility of experimental mating and lack of opportunity (except perhaps rarely in males) to study the meiotic pairing which can be so revealing of homology.

If opportunities for direct test of an interpretation are few, reliance on general hypotheses must be the greater. I make it a rule to doubt an observation if it is not in accord with hypothesis. This is not to reject Bateson's exhortation to “treasure your exceptions”; but one must first make quite sure that they are exceptions. Perhaps one does not make so many new discoveries this way, but fewer may later be shown not to be discoveries after all.