In birds and some other taxa (e.g. urodele amphibians), several spermatozoa typically penetrate the ovum wall at the time of fertilization (referred to as physiological polyspermy) even though the female nucleus fuses with only a single spermatozoan nucleus (Harper 1904, Elinson 1986, Perry 1987). In birds, fertilization takes place in the infundibulum within 15 to 30 min of the ovum being shed from the ovary, at which point one or more spermatozoa penetrate the inner layer of the perivitelline membrane of the ovum (Olsen and Neher 1948). The numbers of spermatozoa in the infundibulum at the time of fertilization can be estimated by counting those trapped on the outer layer of the perivitelline membrane, which is laid down on the ovum soon after fertilization (Gilbert 1971, Wishart 1987). However, the proportion of spermatozoa that undergo the acrosome reaction and penetrate the inner perivitelline layer is not well known (Howarth 1984). The aim of this paper is to describe a simple technique that permits the identification of both: (1) holes caused by the penetration of spermatozoa into the inner perivitelline membrane; and (2) spermatozoa that have not undergone the acrosome reaction and are still attached to the outer perivitelline layer. These two simple techniques open up new possibilities for exploring the dynamics of sperm utilization and sperm competition in birds (Birkhead and M0ller 1992). Methods.-We used fertile and infertile eggs laid within the previous 48 h from naturally mated Japanese Quail (Coturnix japonica). Eggs were opened with scissors and the yolk and albumen separated. The yolk was placed in a petri dish with the germinal disc uppermost. A piece of the perivitelline layer approximately 1 x 1 cm around the germinal disc (diameter = 3.75 mm) was removed with scissors and fine forceps. The piece of membrane was gently washed by shaking in phosphate buffered saline for a few seconds. A second piece of the perivitelline layer (1 x 1 cm) from the same egg, but away from the germinal disc, was also removed and washed. The two pieces were then placed on a microscope slide side by side and laid out flat. We used Wishart's (1987) technique for identifying sperm; this involved placing a drop of fluorescent Hoescht dye 33342 onto the perivitelline tissue and covering with a cover slip. Tissue samples were examined using a transmission microscope. Holes were visible using dark-field optics at a magnification of x 100 (i.e. x 10 objective; Fig. la). Spermatozoa were visible under ultraviolet light at a magnification of x 250 or x 400 ( x 25 or x 40 objective; Fig. lb). The method we used for visualizing holes differs from that of other authors. Bramwell and Howarth (1992b) used Schiff's reagent, and G. J. Wishart (pers. comm.) used a lectin fluorochrome to stain the perivitelline layer in order to increase the contrast between the holes and the layer. However, we found that holes were clearly visible without any stain if we simply used dark-field optics. Moreover, we could use the DNA-specific Hoescht dye 33342 to stain spermatozoa, so that both holes and spermatozoa could be counted from the same piece of perivitelline layer, under dark-field and fluorescence optics, respectively. We examined a single egg laid by each of 15 females, recording the number of holes and spermatozoa over the area of the germinal disc and away from the germinal disc. Counts of holes were made for fields of known area (for holes 3.08 mm2; for spermatozoa 0.48 mm2) and converted to densities per square millimeter. We counted 10 nonoverlapping fields per germinal disc and 30 random fields away from the germinal disc. Densities of spermatozoa and holes were used to estimate the total numbers of spermatozoa and holes over the entire ovum. Each hole in the inner perivitelline layer was assumed to correspond to a single spermatozoa either undergoing the acrosome reaction or undergoing proteolytic activity (see below and Birkhead et al. 1994) and penetrating the layer. Spermatozoa nuclei on the outer perivitelline layer were those that had not undergone the acrosome reaction or proteolytic activity. The surface area (mm2) of the ovum (994 mm2) was calculated using the formula 47rr2 after measuring the diameter of 10 hard-boiled yolks to obtain a mean radius (8.895 mm). Densities are per square millimeter (x ? SE). Results.-The mean densities of both spermatozoa and holes were significantly higher in the germinal disc area than elsewhere on the ovum. The density