staining the same. As the nucleus condenses and elongates to form the sperm head, a light region containing this deeply staining body is seen on one side (figs. 82, 83). A little later the body is divided into two, which appear sometimes spherical (fig. 84), sometimes elongated (fig. 85). As the sperm head elongates still more, approaching maturity, these bodies diminish in size (figs. 86, 87) and ultimately disappear. A cross section of the sperm head at such a stage as figure 87 shows the chromatin in crescent shape with material which stains very little within (fig. 88). The chromatin-like body described above was observed in Tenebrio in a stage corresponding to figure 81, and it was thought that the larger body seen in some cases and the smaller one in others might be the larger and smaller heterochromosomes, but a study of this element in more favorable material disproves that supposition by showing that the different sizes are merely different phases in the evolution of the body. Throughout its history it stains like dense chromatin, and my only suggestion as to its origin is that it seems, from a study of this and other species of beetles, to be a derivative of the chromatin of the spermatid, increasing in size for a time, then decreasing, and finally breaking up into granules and dissolving in the karyolymph. Whether it has any function connected with the development of the spermatozoön, or whether it is merely material rejected from the chromosomes, as in many cases in oögenesis, one can only surmise.

In one testis a peculiar abnormality was found. In all of the perfect spermatogonial plates two small chromosomes were present (figs. 89, 90). Nineteen such plates were counted in five different cysts. All of the equatorial plates of the first spermatocytes showed 8 chromosomes, as usual. In a few favorable growth stages (fig. 91) the two small chromosomes were seen to be combined with the larger heterochromosome and a plasmosome, and one first spermatocyte spindle was found in which the same combination could be clearly seen (fig. 92). All of the second spermatocyte metaphases in which a small chromosome occurred, contained two small ones, making 9 in all (fig. 93). The others contained 8 large chromosomes, as usual. The only explanation suggested by the conditions is that somewhere in its history, the small chromosome had undergone an extra division, and that ever afterward the two products behaved like the one small heterochromosome of a normal individual. The chief interest in this abnormality centers in the fact that the two small chromosomes of this specimen behave exactly like the usual single one, emphasizing the individuality of this particular heterochromosome. Both evidently have the same individual characteristics and affinities as the one in other cases.

Epilachna borealis (Family Coccinellidæ).

Epilachna borealis was found in abundance on squash vines at Woods Hole, Massachusetts, in September. The testes, unlike those of most of the Coleoptera, consist of many free follicles similar to those of the Orthoptera. The germ glands were rather far advanced, but some good spermatogonial and spermatocyte cysts were found. In figure 94, a spermatogonial metaphase, the small chromosome is shown with 17 larger ones. The heterochromosome pair appears in condensed form in the spireme stage (fig. 95), and again in the first maturation spindle (figs. 96, 97). The varying forms of the ordinary chromosomes are shown in figure 98. Figures 99 and 100 are equatorial plates of the first mitosis. The unequal pair is shown by itself in figure 101, and the separation of the heterochromosomes is seen in figure 102. Equatorial plates of the second division, one containing the small chromosome (b), are shown in figure 103. A prophase of the same division (fig. 104) proves that the small chromosome divides quantitatively like the others. It was interesting to find here and there in this material whole cysts in which the nuclei were like those described by Paulmier ('99) for Anasa tristis (plate XIII, fig. 14) as cells which were being transformed to serve as food for the glowing spermatids (figs. 105, 106). The only occasional appearance of these cysts seems to me to preclude their being a special dispensation to furnish the spermatids with nutrition during their transformation. Their appearance and size make me suspect that they are giant spermatids due to the failure of one of the spermatogonial or spermatocyte mitoses. The smaller chromatin body seems to correspond to that described for the spermatids of Odontota dorsalis.

Euphoria inda (Family Scarabæidæ).

Of Euphoria inda only one male was captured, but the numerous testes furnished abundant material in desirable stages. The spermatogonial equatorial plate (fig. 107) contains 20 chromosomes of which the two smallest (l and s) form the unequal pair. The resting spermatogonium contains a two-lobed plasmosome (fig. 108). The growth stages are similar to those in Tenebrio in showing no distinct bouquet stage, but there is a spireme stage in which the heterochromosome pair is clearly seen (fig. 109). Figure 110 (plate XI) is an early prophase, and figure 111 one in which the unequal pair appears with a tetrad and several dumb-bell forms. The prophase of the spindle, as in Odontota, is much elongated (fig. 112). In figures 113-116 the small heterochromosome pair is shown in various positions with reference to the other chromosomes of the metaphase of the first spermatocyte. Figure 117 shows it more deeply stained than the others in the equatorial plate. This pair divides in advance of the others, and the larger and smaller elements are plainly seen nearer the poles in anaphase than the other univalent chromosomes (figs. 118-120). Daughter plates of the first spermatocyte are shown in figure 121, and equatorial plates of the second spermatocyte in figure 123. Figure 122 shows the telophase of the first division with the spindle for the second division forming. In figures 124 and 125 we have daughter plates of the two classes of second spermatocytes, showing the content of the two equal classes of dimorphic spermatozoa, as this was shown in Tenebrio. Figures 126 and 127 are anaphases showing the division of the heterochromosomes (l and s). Figures 128-130 are early stages in the development of the spermatid showing the chromatin nucleolus (n) in various phases.

Blepharida rhois (Family Chrysomelidæ).

The testes were rather too far advanced when this material was collected and no dividing spermatogonia were present. The growth stages (figs. 131, 132) show a faintly staining spireme and a heterochromosome group similar to that of Odontota, a large and a small chromosome attached to a large plasmosome. The spireme appears to go directly over by condensation and segmentation into the dumb-bell-shaped figures seen in the first maturation spindle (figs. 133, 134), though cross-shaped bivalents occasionally occur (fig. 135). The heterochromosome pair, slightly separated by plasmosome material, is usually found at the periphery of the plate (figs. 133-136). Figure 137 is an exceptional anaphase in which the heterochromosome elements are not mingled with the polar masses of chromatin. Figures 138 a and b are equatorial plates of the second mitosis, and figures 139 and 140 are pairs of daughter plates from second spermatocytes showing again the dimorphism of the spermatozoa as to their chromatin content. As in several of the forms studied, material was collected for examination of the somatic cells, but no favorable cases of mitosis were to be found.

Silpha americana (Family Silphidæ).

Only one male of