Aphis œnotherae.
The spermatogenesis of Aphis has been fully described in another paper and will merely be briefly summarized here for the purpose of comparison with other forms.
The spermatogonia contain a large nucleolus, which gradually disappears in the prophases of mitosis (plate VII, figs. 209-211). The youngest spermatocytes closely resemble the spermatogonia (fig. 212). There is no bouquet stage and no such marked spireme stage as in many other insects. The true synapsis occurs, as shown in figure 213, by pairing of like chromosomes side by side. This conjugation of like chromosomes is followed by a stage in which they are massed together at one side of the nucleus (fig. 214). In these latter stages the nucleolus has entirely faded out and nothing suggesting an accessory chromosome is present. Figures 215 and 216 are equatorial plates of the first spermatocyte mitosis. There are 5 chromosomes of different sizes and shapes, and figure 216 shows each one double. The first division of the chromosomes, though apparently longitudinal, is evidently a separation of the elements paired in a preceding stage, and is therefore a reducing division.
The anaphase of the same mitosis is shown in figures 217 and 218; it is peculiar in that one chromosome always divides more slowly than the others, the two elements hanging together at one end. In figure 219 are sister spermatocytes of the second order, the "lagging" chromosomes still connected. The second maturation division is seen in metaphase in figure 220 and in anaphase in figure 221. Figure 222 shows a young spermatid, the five chromosomes still preserving their characteristic form. Figure 223 is the equatorial plate of the first maturation division of the winter egg, showing the same form and size relations of the chromosomes as in the spermatocyte divisions. Figures 224 and 225 are equatorial plates of a polar spindle (fig. 224) and of a segmentation spindle (fig. 225) of the parthenogenetic egg, where 10 chromosomes are present, 2 of each of the sizes found in the sexual germ cells.
So far as an accessory chromosome or any other visible evidence of a sex determinant are concerned, the results are entirely negative. The conditions shown do, however, support Mendel's conception of the "purity of the germ-cells," and also afford evidence in favor of Boveri's theory of the individuality of the chromosomes.