The relative merits of the various biological tests for pregnancy.

Among the many purely magical methods of pregnancy diagnosis practised by the Ancient Egyptians, were tests based on the germination of seeds and rate of plant growth when treated with pregnancy urine (Bayon, I939). That there is some scientific basis for these tests is shown by the general agreement that a factor contained in pregnancy exerts a beneficial effect on plant growth. It is of interest to note that Orban (i947) (cited by Cowie, 1948, p. 48), has described a test depending on the rate of opening of gladioli blooms when placed in pregnancy urine. Modern tests date from Aschheim and Zondek's (I927) observation of the effects of chorionic gonadotrophin on the ovaries of immature mice, and to their subsequent application of this observation as a basis for pregnancy diagnosis. It is difficult to present a balanced, yet critical, account of the many tests evolved since that time. The use of a particular test in one's laboratory implies a certain bias in itsfavour. For this reason it is important to have a clear conception of the desirable features of the ideal pregnancy test, and with reference to this to see how much and in what ways tests fall short of the ideal. As defined in the Lancet Editorial (I948): 'Ideally a pregnancy test should be reliable, convenient, simple and speedy; and those involving the patient's participation, fall short of the ideal if they are irksome or require her to visit anyone but her own doctor.' On the grounds of reliability we may exclude many of the more exotic tests. Thus the development of nuptual colouration in the Japanese Bitterling as a positive response is too irregular for use in pregnancy diagnosis (Owen, 1936). Tests such as the Konsuloff test based on the chromatophore hormone and its many modifications can similarly be condemned. The chromatophore hormone is present in the blood and urine of normal men and women, but is present in increased quantities during pregnancy. Konsuloff (I934b) using hypophysectomized frogs showed that darkening of the skin caused by dispersal of the pigment granules in dermal melanophores occurred in I-4 hours following injection of pregnancy urine. He claimed an accuracy equal to that of the Aschheim-Zondek test.