SEX ATTRACTANTS IN PRIMATES 259 now been applied to the sexual skin area of ovariectomized recipient rhesus monkeys on 249 separate occasions, and a 1 h behaviour test observed on each occasion in an acoustically isolated testing booth housed behind a one-way screen. These tests involved 11 pairs of animals and the secretions significantly increased the male partners' sexual behaviour above pre- treatment periods when only control substances were applied (Fig. 1). In Pre - treatment Treatment Pre-treatment Treatment 12 - 0.6 - PO.O 200 !49 • 0.4- .•, • 0-2- / PO-001 / 20O 249 Figure l. Effects of applying vaginal secretions from oestrogenized 'donor' females to the sexual skin area of ovariectomized 'recipient' on the sexual stimulation of males. Data for 11 pairs involving five males, five females and five donors. the pre-treatment period eight ejaculations were recorded in 200 tests and these increased to 139 during 249 tests when vaginal secretions were smeared on the sexual skin of unreceptive recipients (C •'= 96.04, P 0.001). Of even greater significance was the increase in male mounting attempts from 174 during the pre-treatment period to 2292 during applications to the same female partners. This high number of mounting attempts with few ejacula- tions was an indication of the unreceptive condition of the ovariectomized recipients, and clearly demonstrates the males' increased sexual interest in these pheromone treated females. To determine the chemical nature of the substances in vaginal secretions responsible for these powerful behavioural effects, extraction and fractiona- tion procedures were used in conjunction with behavioural assay methods.

260 JOURNAL OF THE SOCIETY OF COSMETIC CHEMISTS The early stages of this procedure involved the use of ether extracts of secretions, collected by lavage with water from oestrogen-treated donor females (8). Ovariectomized rhesus monkeys were again the recipients for these extracts and the very low levels of sexual activity during the pre- treatment period was in marked contrast to the high levels seen during the applications of ether extracts. A gas chromatographic comparison of ether extracts of the vaginal secretion from ovariectomized untreated females indicated that the amounts of volatile components were absent or low, while oestrogen treatment stimulated production of volatile components and improved the sex attractant properties of vaginal secretions (11). Identification of these volatile components was obtained by preparative gas chromatography and mass spectrometry (5). The resultant mass spectra were compared with authentic samples, and established the identification of the first five peaks as acetic, proprionic, isobutyric, butyric and isovaleric acids. A mixture of authentic acids was made up to match their concentra- tion in a pool of vaginal washings, and a small sample of this mixture when tested for behavioural activity was demonstrated to possess sex attractant properties (Fig. 2). The effectiveness of these pheromones in stimulating sexual behaviour in the rhesus monkey does, however, vary according to social conditions: with some partners and in certain tests no sexual stimulation occurs. As more males are tested it is becoming evident that the response to pheromones varies between individuals and is also dependent in part upon the female partner with which they are paired. When the behavioural effects of fresh vaginal secretions and a synthetic mixture of their acid content were com- pared in the same nine pairs of animals, vaginal secretions appeared to be more effective in stimulating the male's sexual behaviour (Fig. 3). Although both vaginal secretions and the synthetic pheromone complex stimulated male sexual activity at significantly higher levels than in the pre-treatment tests, vaginal secretions were effective in 59•o of applications compared with only 35•o effectiveness of the synthetic acid mixture. Moreover, vaginal secretions stimulated 452 male mounting attempts compared with only 257 during application of synthetic pheromone to the same female partners (t= 2.39 P 0.02). This lower proportion of effective tests during applications of the synthetic pheromone is due to its failure to stimulate mounting behaviour in certain pairs (Fig. 4). The relative effectiveness of untreated vaginal secretions and the synthetic acid mixture varied with the male partners from 100•o in the case of male 113 to 45•o success with male 68. With males 113