Sperm develop in the seminiferous tubules of the testis, beginning as stem cells, spermatogonia. Some of these undergo a final replication of their DNA and begin a protracted prophase of the first meiotic division, becoming primary spermatocytes. After completion of that division, the resulting secondary spermatocytes promptly undergo the second meiotic division, to become spermatids. The differentiation of spermatids involves production of flagellum and acrosome, shaping of the head, condensation of nuclear chromatin, formation of mitochondrial sheath and various supportive structures of the flagellum, and loss of excess cytoplasm. Mature spermatids are released into the lumen of the seminiferous tubule, within which they journey to the epididymis, where they will complete their functional differentiation into mature sperm.
Sperm development (spermatogenesis) is a continuous process. However, if you examine a histological section of rat testis, you will not see the complete process represented in any particular seminiferous tubule. Instead, each seminiferous tubule cross section contains 4 or 5 distinct phases of spermatogenesis that were developing in tandem at that level in the seminiferous tubule when it was fixed for light microscopy. You will note that the set of spermatogenetic phases appears to be different in each tubule cross section. If you had infinite time and patience, and were willing to photograph the whole section and cut out a picture of each tubule, you could arrange the tubules in a series from left to right in which the germ cells in each of the 4-5 layers of the successive tubules were progressively more developed. It would then be possible for you to trace the complete process of spermatogenesis by starting with spermatogonia in the first tubule's lowest layer and proceeding to the right to the end of the series, then go back to the beginning tubule and continue with the second layer, and so forth until you reached the top layer, where mature spermatids would be released. Fortunately, you don't have to go through all of this, since the sequence of these tubule cross sections has already been worked out, and is called the cycle of the seminiferous epithelium. For convenience, the possible tubule cross sections in this cycle have been divided by Clermont and Leblond into 14 stages, which are shown in the diagram below (from Perey, Clermont and Leblond 1961). For a bibliography and brief history of the cycle of the seminiferous epithelium, click here.
Unless otherwise stated, the photos show rat testis fixed in Bouin's fixative, embedded in paraffin, sectioned at 6 µm thickness, and stained with PAS-Weigert iron hematoxylin (histological slide from Dr. Philip E. Smith).
Diagram and table showing the cycle of the seminiferous epithelium in the mouse (from Oakberg 1956, American Journal of Anatomy 99:391-413).
A poorly-developed cycle in the human seminiferous tubule occurs in patches and is not easy to discern in histological sections (from Clermont 1963, American Journal of Anatomy 112:35-51).