In The Female Sex Hormone (1929) Robert Frank suggested: “The female principle which we call the female sex hormone is widely distributed throughout the vegetable and animal kingdom.” Frank described the presence of the “female principle” in a wide variety of substances, not only originating from animals and not even restricted to the females of the species. The list in this first handbook on female sex hormones included substances such as yeast, the buds of willows, potatoes, sugar beet, rice, ovaries and placenta, the body fluids of males such as blood, urine and bile, and even testes (Frank 1929: 109).
How did scientists decide which substances they would classify as female sex hormones? In the early decades of the twentieth century scientists suggested that the label “female” should be attached to substances that could restore the organism after removal of the ovaries. This criterion enabled researchers to choose from a whole series of suitable testing methods. In theory, any character or function that changed in the organism after the removal of the ovaries, and could be restored to its original state by the administration of preparations which were expected to contain the “female principle,” could serve the purpose of a bio-assay for female sex hormone (Frank 1929:887). Prior research on the removal of the ovaries had already provided the requisite knowledge as to which functions or characteristics were affected by ovariectomy.
In the early years scientists largely agreed which substances were suitable to work with. Following the paradigm which considered the ovaries as the seat of femininity, they used extracts prepared from ovaries. There existed, however, less agreement about the test methods. Analysis of textbooks and journals of the 1920s shows that scientists applied a wide range of biological assays in deciding which substances they would label as female sex hormones. Each discipline seems to have adopted its own specific test. In the early studies, gynecologists in Paris and Vienna used the growth of the uterus —or more precisely the increased weight of the uterus in ovariectomized rabbits after administration of ovarian extracts—as a decisive criterion for the labeling of substances as female sex hormones (Gustavson 1939:887). The choice of this test can be understood in terms of the tradition in the life sciences of considering the uterus as the female organ par excellence. Moreover, gynecologists preferred this test because of their professional interest in pregnancy. In addition to using this biological assay method, gynecologists also performed clinical trials in which ovarian preparations were evaluated for use in the treatment of various kinds of diseases associated with dysfunction of the ovaries.
In contrast to gynecologists, laboratory scientists introduced test methods based on a wider variety of functions and characteristics. Physiologists and zoologists tested ovarian preparations on the feathers of domestic fowl; the growth of mammary rudiments in male mice; the growth of the vulva and the mammary glands in infantile female rats; and muscular activity, basal metabolism, and the levels of calcium and sugar in blood, both in mice and in women (Frank 1929:109). The fact that physiologists opted for a wider variety of tests than did gynecologists can be understood in the context of the tradition of then — discipline, which studied the physiology of the animal organism in all its aspects. The different tests served different purposes. The feathers test, for instance, was used primarily to study the duration of the action of ovarian preparations. This bio-assay was considered a rather convenient test, because the administration of ovarian extracts left a continuous and permanent record on the feathers instead of merely being reflected in the status of their growth at a particular time, as was the case in the uterus test (Parkes 1985:515). The test involving the growth of mammary rudiments in male mice was used as an indicator for very small amounts of ovarian preparations (Allen et al. 1939:515).
Behavioral tests were also introduced, particularly by psychologists involved in the study of sex hormones. William C. Young, an animal psychologist at the University of Chicago, used reproductive behavior as a specific test for female sex hormones. Studies of reproductive behavior in animals had indicated that among the lower mammals, which include species frequently used in the laboratory such as guinea-pigs, rats and mice, the females usually mate only during a specific period, called estrus. Experiments indicated that these periods of sexual activity disappeared after ovariectomy, but could be restored by injections of ovarian preparations (Allen et al. 1939:503). This behavioral test procedure was applied only in the early studies of sex hormones.1
The laboratory scientists and the gynecologists, however, could not agree about which test to accept as a decisive proof that an ovarian extract contained the “female principle.” The laboratory scientists criticized the gynecologists for testing ovarian preparations in the clinic before their physiological and pharmacological effects had been determined. The biological test method developed by the gynecologists (the uterus test) was also criticized by laboratory scientists as inadequate (Borell 1985).2 This dispute between gynecologists and laboratory scientists took place in the context of a growing need on the part of the pharmaceutical industry to improve the quality of its ovarian products. Gynecologists had claimed that the powders and pills available from the drug trade did not contain any active ovarian substance. In 1929 Frank warned his colleagues of the poor quality of these products:
commercial extracts now placed upon the market show a woeful lack of potency and rapid deterioration. Unpleasant local reactions may arise at the site of injection. The prices of these pharmaceutical preparations are prohibitive. Consequently I warn against their general use until better products are at our disposal.
(Frank 1929:297)
Actually, skepticism about the quality of commercial ovarian products was the very reason why many gynecologists (e. g., the German gynecologist Bernhard Zondek) became involved in research on ovarian preparations.3 This criticism emanating from the gynecological community was part of a growing professional concern about the quality of all types of drugs (Starr 1982:131-132). Consequently, the pharmaceutical industry also became interested in biological test procedures that could confirm the quality of its products.
The dispute over the appropriate methods of assaying sex hormones was part of a more general struggle between laboratory scientists and clinicians, which can be seen as characteristic of this period in medical history. The early decades of the twentieth century were marked by the growing professionalization of the sciences, a process in which laboratory scientists presented themselves as the dominant professionals among those, including clinicians, who were concerned with natural phenomena. By emphasizing the superiority of physiological methods over therapeutic test methods, laboratory scientists claimed the position of experts on tests, thus defining the demarcation lines of their own profession (Borell 1985: 4,11,18).
In 1923, laboratory scientists introduced a new test method that in their opinion was much better than the prevailing methods for evaluating ovarian preparations. Two American scientists, Edgar Allen, a professor of anatomy at the Yale University School of Medicine, and Edward Doisy, a professor of biochemistry at the Medical School of the University of St Louis, introduced the vaginal smear test, which was based on cyclic changes in the epithelial cells of the vagina considered characteristic of estrus. In the 1920s, some gynecologists and pharmaceutical manufacturers started to apply this test in their research.
Laboratory scientists used other strategies as well to establish their position as experts in the growing field of sex endocrinology. Even more important than introducing new tests was their role in setting standards to reduce complexities in a field that was becoming increasingly complex. By the end of the 1920s, there existed a bewildering variety of test methods, since every new group that became involved in research on sex hormones introduced new assays. Laboratory scientists tried to handle these complexities by setting standards for tests and terminology. The need for standardization in general was strengthened by the difficulties that had been experienced by physicians, the public health services, and pharmaceutical manufacturers due to the existence of a variety of different units for expressing the potency of one and the same drug. The situation had been particularly troublesome during the First World War. As a result, the Health Organisation of the League of Nations initiated a series of international conferences to establish international agreement on biological standardization in various fields, including the research and development of sex hormones (Gautier 1935). British laboratory scientists played a major role in this process of standardization. The National Institute for Medical Research in London had at its disposal a special Department of Biological Standards, charged with establishing international biological standards for drugs, hormones and vitamins (Hartley 1936). Two conferences were organized for the standardization of sex hormones, both chaired by Henry Dale, the director of the National Institute for Medical Research. These conferences provided the first occasion for international discussion in the emerging field of sex hormones research. Physiologists and biochemists from Britain, the United States, France, Germany, Switzerland and The Netherlands participated in the standardization conferences and came to an agreement on standard tests, a common terminology and the development of biological standards (Dale 1935 a; 1935b).
The first Conference on the Standardization of Sex Hormones, in 1932, settled the question of which biological test was most appropriate for labeling substances as female sex hormones. The vaginal smear test was accepted as the standard test for these substances, which were now called “oestrus — producing hormones.” The conference agreed
that by the term “specific oestrus-producing activity” is to be understood the power of producing, in the adult female animal completely deprived of its ovaries, an accurately recognizable degree of the changes characteristic of normal oestrus. For the present, the only such change regarded by the Conference as providing a suitable basis for quantitative determination of activity in comparison with the standard preparation is the series of changes in the cellular contents of the vaginal secretion of the rat or mouse.
(Dale 1935a:124)
In this manner, laboratory scientists reduced the complexities of hormonal research. The choice of one specific test method as a standard test simplified their experimental practice. The use of a standard test enabled scientists to compare their experimental data with their colleagues’ experiments. At the same time, the selection of one specific test also restricted the broad scope of hormonal research. In 1936, Dutch scientists criticized what they called the “unitary school” in sex endocrinology, suggesting that the search for suitable test methods led scientists to neglect the other biological functions of sex hormones.4
In this process of standardization, the laboratory scientists, and not the gynecologists, determined which test was accepted as the standard test for female sex hormones. With the development of the vaginal smear test, laboratory scientists provided the other groups in the field with the techniques required to improve the quality of hormone preparations. Dutch laboratory scientists evaluated the introduction of the vaginal smear test as follows: “The work of Allen and Doisy enables us to replace vague concepts and general reflections by measure and number” (E. Laqueur et al. 1927a:2, 077). The activity of ovarian preparations was expressed in animal units, referring to the quantity required to restore estrus changes in the vagina of ovariectomized animals, a practice which showed a remarkable difference in terminology between Europe and the USA. Because European scientists usually experimented with mice and American scientists worked only with rats, “the female principle” was expressed in mouse or rat units, according to which side of the ocean one was on (Dale 1935b).
Although the first standardization conference reduced some of the complexities of the field, scientists soon faced other complexities. We saw in Chapter 2 how scientists initially opted for the existence of just one single female sex hormone. It was in the process of testing substances isolated from the ovaries that scientists gradually abandoned the “one hormone” doctrine. This meant that scientists had to decide again about standardization problems. Three years after the first conference took place, the Second Conference on the Standardization of Sex Hormones (1935) was organized to discuss the terminology and standard tests for the “second” female sex hormone. At this conference scientists accepted the name “progesterone” for common use in scientific literature. The test on “the proliferation of the endometrium (uterus tissue) of the rabbit,” first introduced by the British anatomist George Corner, became accepted as the standard test for progesterone (Dale 1935b).
In the 1910s and 1920s, the measuring of female sex hormones was solely in the hands of biologists and gynecologists. This situation changed drastically in the late 1920s. As a result of the introduction of urine as a research material, chemical methods of analysis came to be of greater importance, and consequently the new discipline of biochemistry became more and more involved in research concerning sex hormones.5 Biochemists devoted their energies to the chemical isolation and identification of sex hormones.
This change in focus from biological to chemical methods was part of a general emphasis on chemical determinations in the process of standardization. The chairman of the standardization conferences on sex hormones, Sir Henry Dale, expressed this preference for chemical methods in 1935 as follows:
It may be further recognised that biological methods of comparative assay, with their inherent limitations of accuracy and their heavy cost in material and skill, should be replaced by chemical determinations, whenever these can be made with as great accuracy and specificity as those based on animal tests. It may be hoped that, as knowledge of the biological important principles increases, remedies which still, with our present knowledge, can only be standardised by biological methods may pass, one by one, into the class in which chemical analysis will supersede biological standardisation.
(Dale 1935b:621)
After the chemical isolation and identification of the sex hormones in the 1930s, the labeling of substances as female sex hormones was no longer defined primarily by biological tests. Researchers were now able to attach the label “female” to ovarian and urine preparations following the analysis of their chemical structures. In the early 1930s, Laqueur’s research group was the first to apply a chemical test procedure for female sex hormones, consisting of a color reaction for the quantitative determination of these substances (Walsh 1985:12). Chemical test procedures, however, did not completely replace biological tests. Biological assays remained in use as devices to control the quality of chemically manufactured hormone preparations (J. Freud 1938:1,667).