The studies described in this thesis have been concerned with several aspects of follicular development in the mammalian ovary. Chapters 2, 3 6 4 deal with mathematical modelling of ovarian follicle dynamics in normal animals and comparisons with experimentally manipulated animals. Chapter 5 describes a novel aethod for estimating the clonal origin of the mouse ovarian follicle. In the final two chapters, the comparative physiology and anatomy of follicular numbers and sizes and the incidence of polyovular follicles are described for a number of species. The unifying theme of these studies is that they reveal patterns existing in follicular development and utllisation by detailed examination of one species, (CBA/ca mouse), and broadly by interspeci,fic comparisons (with relation to scaling). A detailed mathematical description of the follicular dynamics of virgin CBA/ca mice up to 98 days of age has been obtained by the application of compartmental modelling to differential follicle counts. The rates of follicle growth (migration) and death have been estimated for five ?stages of development (primordial to Graafian). The model predicts age changes in follicle growth and death rate, there being transitions in the parameters at 20 days second at 60 days. The parameters for normal animals have been compared with those of anilllals under two experimental conditions: 1) by unilateral ovariectomy at 4-2 days of age, which abruptly halves the numbers of ovarian follicles and alters the ratio of large : small follicles. 2) by blocking ovulation using progesterone implants. The dynamics of follicle growth were altered by both treatments in comparison with the controls. Follicles at all stages of development were affected by unilateral ovariectomy and differences may exist with time. The compensatory response by the remaining ovary was due to a combination of an increased preantral growth rate and a decrease in atresia at antral stages. Earlier stages of follicle development were affected this may have been incidental to the compensatory response. In progesterone treated animals follicles developed through to antral stages when they un~erwent atresia. The effects of treatment were observed at three levels of development: 1) The initiation of growth from the primordial pool, 2) Growth rate of small follicles and 3) deaths at larger stages of follicular development. Longer term observations indicated that these effects may not be constant. The modelling studies have looked at numerical changes in the follicle population with time but a greater understanding of the develomental biology of the follicle is required in order to explain the changes in growth and death rates observed. This problem has been tackled initially by studying the clonal origin of the follicular epithelium. The technique used is based on the principle that cells in females. are generally mosaic as a result of X-chromosome inactivation the use of X linked cell markers phospho-glycerate kinase-1 (PGK-1). Granulosa cells were found to be polyclonal in origin with the number of progenitor cells numbering 5 on average. Analysis of cumulus and mural granulosa cells showed that substantial cell mixing had occurred and cuaulus cells were generally founded by more than one clone. Finally, comparative studies have been conducted to look at scaling of follicle sizes and numbers and of polyovular follicles. Ovarian follicle and oocyte sizes were scaled according to body weight (ranging from .005-500Kg) using data from 22 species. Primordial and Graafian follicle sizes varied with body weight but closer correlations for the latter were obtained when the sum of the surface areas or volUiles for a preovulatory set were considered as opposed to the values for individual follicles. The numbers of nongrowtng follicles 1n reserve at young adult ages were correlated with maximum longevity of the species and related to body weight. The frequency of polyov~lar follicles varied 1n species studied and were most abundant 1n the domestic bitch. The overall incidence of polyovular folUcles 1n young bitches was 14 S, being reduced to 5~ 1n bitches at 7-11 years. The frequency of the various types of polyovular preantral folUcle varied inversely with the numbers of oocytes per follicle.