Quantitative and qualitative changes in the follicular apparatus of the guinea pig during the estrous cycle have been extensively studied (Meyers, et al., 1936; Dempsey, 1937; Hermreck and Greenwald, 1964) . In contrast, save for the reports of Loeb (1911), Dempsey (1937) and Rowlands (1956) reporting vesicular follicles in the pregnant guinea pig comparable in size to those found during the estrous cycle, no extensive quantitative data on changes in the ovary at various stages of pregnancy are available for this frequently used species. Moreover, while earlier workers (Dempsey, 1937; Boling and Hamilton, 1939) studied effects of gonadal steroid hormones on follicular growth, they only reported on the size of the largest few follicles. In the present study we have attempted to extend the basic ovarian data available for the guinea pig and have analysed the total follicular count and distribution in cyclic, pregnant, postpartum, progesterone, and testosterone propionate treated animals. In addition concomitant analysis was made of changes in the ovarian weight during these various reproductive conditions. Preliminary reports have appeared (Labhsetwar and Diamond, 1965a,b) .
MATERIALS AND METHODS
Guinea pigs of heterogeneous stock with established regular cycles (16-18 days) were used for this study. Most of the animals were nulliparous. The estrous cycle and pregnancy were dated from the day when vaginal membrane was fully open (Day 1) . A total of 10 groups of animals (4-7 animals per group) was studied. The cyclic animals were killed on Days 1, 6, 10-11 (hereafter called Day 11), and 15 of the cycle while the pregnant animals were killed at the ends of each of 3 trimesters of gestation; Day 21-22 (hereafter called Day 22), Day 44, and Day 65-66 (hereafter called Day 66) . The normal length of gestation in this stock of guinea pigs is 68 ± 2 days (SE) . In addition, the nursing animals were killed on postpartum Day 3 (day of parturition, Day 1) .
Two steroid treatments were investigated. One group of females was injected with progesterone (10 mg/day in 0.2 ml corn oil, subcutaneously) for 16 days beginning Day 2 of the cycle. The day of autopsy would correspond to Day 1 of the subsequent 18-day cycle. The second group received an initial injection of 5 mg of testosterone propionate (0.02 ml oil, intramuscularly) on Day 2 of the cycle, followed by 1 mg daily for 48 days. All the steroid-treated animals were killed a day following the last injection.
The animals were killed by an intracardiac injection of overdose of Nembutal. The ovaries were weighed and fixed in 10% Formalin. Any animal displaying, at autopsy, any gross ovarian abnormality such as cystic condition was discarded.
A pair of ovaries from each animal was embedded in paraffin, serially sectioned (15μ),and stained with hematoxylin-eosin. Every 6th section was projected at a constant magnification (x20), and all vesicular follicles with apparently normal ova were plotted on a paper. Follicles of the first two stages of atresia as cited by Williams (1956) were also included in the count. Subsequently, using a semiautomatic particle size analyzer (Zeiss) the plotted follicles were classified into different groups according to their largest diameter: <300, 301-500, 501-700, and >700μ. This method of plotting and automatic size analysis is accurate and reproducible.
No statistically significant changes in the ovarian weight were observed during the estrous cycle or the first trimester of pregnancy (Fig. 1) . During the 2nd and 3rd trimester, however, the ovarian weights increased markedly (p < 0.01) but returned to a cyclic level by postpartum Day 3 (Fig. 1) .
Treatment with progesterone did not result in a significant decrease in the ovarian weight when compared to Day 1 of the cycle, but treatment with testosterone propionate did (Fig. 1) .
Both the absolute (number/animal, Fig. 1) and the relative (number/ 100 mg ovaries) number of vesicular follicles remained more or less constant (p > 0.05) during the cycle, but during pregnancy and postpartum period the follicular population declined significantly (p < 0.01, Fig. 1) . The mean number of follicles in cyclic animals was 101 ± 9(SE) while a similar mean for the pregnant animals was 67 ± 5 (p < 0.01) .
Both steroid treatment failed to affect the follicular population significantly when compared to Day 1 of the estrous cycle. The ovaries of the progesterone-treated animals in fact contained slightly more follicles than the ovaries of animals killed during the cycle (117 vs. 101, p > 0.05) .
Distribution of follicles in different size categories is shown in Fig. 2. In general the ovaries during the follicular phase (Days 11 and 15) of the cycle contained proportionately more larger-sized follicles than the ovaries during the luteal phase of the cycle (Days 1 and 6) . During pregnancy, although follicular population declined, the follicular distribution was comparable to that found during the luteal phase of the cycle. The follicles larger than 700μ in diameter were found in a majority of animals during preovulatory period (on Day 15 of cycle in 3 out of 5 animals) and terminal stages of gestation (on Day 66 in 6 out of 6 animals) but were rarely encountered at other stages of the cycle or pregnancy. Therefore they can be regarded as preovulatory. Such follicles were never found in postpartum or in any of the steroid-treated animals.
The progesterone treatment failed to inhibit follicular development except for preovulatory growth (> 700μ) . In fact the follicular distribution in the progesterone-treated animals was remarkably similar to that seen in the luteal phase of the cycle. In contrast, the androgen treatment decreased follicular growth considerably since over 99% of the follicles in these animals were smaller than 500μ in diameter (Fig. 2).
The size of corpora (estimated by multiplying the two largest diameters) increased to a maximum by Day 11 and then decreased by Day 15 of the cycle. In the pregnant animals some increase beyond the maximum encountered during the cycle was noted, confirming the observations of Rowlands (1956) . Fresh corpora with blood-filled cavities were present in the postpartum animals while no corpora were found in either of the steroid-treated group. On Day 1 of the cycle some animals had ovulated while others had not and their ovaries contained degenerating luteal tissue from the previous cycle.
In general our findings for ovaries during the estrous cycle are similar to those reported previously by Hermreck and Greenwald (1964) . Some noticeable differences, however, were seen. While we both found the fewest and greatest number of follicles on Day 6 and Day 11, respectively, our values for Day 11 and Day 15 are considerably lower. The reason for this is not presently clear. However, considering the differences in techniques employed and rapid fluctuation in follicle count which seems apparent, the disparities between the findings reported here and those given previously may not be significant.
Our findings in regard to follicle growth during pregnancy in the guinea pig are new. Available information indicates that follicular growth continues during pregnancy in species with a short gestation period such as the rat (Greenwald, 1966; Schwartz and Talley, 1968), hamster (Greenwald, 1964), and mouse (Greenwald and Chaudary, 1969) while in species with a long gestation period such as the cow (Nalbandov and Casida, 1940), sow (Nalbandov, 1953), and ewe (Williams et al., 1956) follicular growth declines. The present study in the guinea pig, a “medium” gestation-length animal (68 days), suggests that while the total follicular population declines during pregnancy, selected follicles continue to grow. Almost invariably several follicles attain a preovulatory size of 700μ or more by Day 66 of pregnancy. While postpartum ovulation is nonexistent or quite rare in the long-term gestational farm animals such as the cow, sow, or ewe (see Cole and Cupps, 1959), in the guinea pig postpartum ovulation is the rule. Three days after parturition practically all ovaries contained fresh corpora lutea; follicles larger than 500μ in diameter were rarely seen. This is a situation comparable to that seen on Day 6 of the cycle. Apparently preovulatory follicles found on Day 66 of pregnancy had ovulated.
The number of follicles significantly decreased during pregnancy. Since follicles during this period remain responsive to exogenous HCG (Rowlands, 1956), this selective decline in the follicular population probably cannot be attributed to a decrease in ovarian sensitivity to gonadotropins. It is possible that rising levels of progesterone (Heap and Deanesly, 1966) during pregnancy provide a feedback mechanism working to decrease the output of gonadotropins from the pituitary and thus account for the decline in follicular number. It is also possible that ovarian or placental steroids cause qualitative changes in gonadotropin secretion, such as an alteration in the FSH/LH ratio, which favors an increased rate of atresia resulting in the decrease in follicular number.
Since atretic follicles have been considered a possible source of interstitial tissue (Harrison, 1962), an increased incidence of atresia during pregnancy may contribute towards an increase in interstitial tissue. While Stafford and Mossman (1945), without recording ovarian weights, reported no correlation between interstitial tissue and pregnant state in the guinea pig, we feel the increase of interstitial tissue, cellular and fluid, might be crucial in accounting for the significant ovarian weight increase during the second and third trimesters of pregnancy.
In an early study Dempsey (1937) reported that a dose of progesterone effective in inhibiting ovulation failed to decrease the size of the largest 3-4 follicles in the ovaries. Deanesly (1968), utilizing progesterone implants to effect a minimum of 0.1 mg absorption daily, similarly reported inhibition of ovulation. Without presenting actual follicle measurements, in the absence of which direct comparison is difficult, she noted normal follicular growth and regression in these animals. The present study, employing considerably larger doses than those used by either of the previous investigators and involving measurement of the total follicular population in the ovaries, indicates that even relatively large doses of progesterone do not appreciably affect follicular size if comparison is made with Day 6 ovaries. Comparison with ovaries of other cycle stages, however, indicates that the growth of the larger size follicles (preovulatory follicles) was indeed inhibited (i.e., that growth which required synergism of FSH with LH) . In this respect the guinea pig resembles the rat (Rothchild, 1965) and hamster (Labhsetwar, unpublished data) where progesterone treatment has also been found to inhibit only the preovulatory growth of follicles.
Prolonged administration of testosterone propionate was quite detrimental to the ovary. Testosterone propionate inhibited ovulation, seriously inhibited follicular growth, and caused a significant decrease in ovarian weight. These results reaffirm earlier findings (Boling and Hamilton, 1939; Deanesly, 1968).
While a modicum of progress has been made in exposing our unknowns in ovarian functional morphology for this widely used species, many of the basic questions regarding the nature of the atretic process or the ovarian gain in weight remain unanswered (see Mossman, 1968) .
This basic study was supported by grant GM-10632 from the National Institute of Health, and grant GU-786 from the National Science Foundation while the authors were members of the Department of Anatomy, University of Louisville School of Medicine. The research was completed with partial support from the Ford Foundation and grant HD 03394 of the NIH USPHS. Grateful acknowledgment is also made to Lederle Pharmaceutical Co. for a Lederle Medical Faculty Award to one of us (MD) . Finally, technical assistance of Mrs. Barbara Tullis is appreciated.