equal classes of spermatozoa, one class containing 11 ordinary chromosomes, the other 11 ordinary chromosomes plus the odd heterochromosome, for the odd chromosome divides with the others in the second spindle as in Orthoptera (McClung and Sutton).

In figures 264 and 265 (plate XV) are seen the telophase of the two kinds of second spermatocytes, one (fig. 265) showing the divided odd chromosome, which continues to stain more deeply after the others have become diffuse. All of the spermatids (figs. 266-268) contain, in the early stages of development, a body (n) which stains like chromatin, but increases in size from a small granule in the telophase (figs. 264, 265) to the large dense body (n) seen in figure 267. This is probably homologous with the chromatin nucleolus described for the spermatids of the Coleoptera. In addition to this, in one-half of the spermatid nuclei there is a condensed mass of chromatin which is evidently the derivative of the odd chromosome of the spermatogonia and spermatocytes (figs. 267 and 268, x). In common with the spermatids of other Hemiptera these show two masses of archoplasm, the larger of which forms the sheath (s) of the axial fiber of the tail, and the smaller the acrosome (a). The axial fiber grows out directly from the centrosome, on either side of which there is a dense band forming the lateral boundary of the middle piece. It will be seen that the odd chromosome of Aphrophora is in its behavior precisely like the typical Orthopteran "accessory" of McClung, and similar to the odd chromosome of the Coleoptera.

In various parts of the young male larvæ dividing cells were found and the number 23 determined (fig. 269). Turning now to the female larvæ to determine the somatic number, the oögonia proved to be more favorable for counting. Twenty-four chromosomes were present in equatorial plates of oögonial mitoses (fig. 270), thus confirming Wilson's results for the Anasa group of the Hemiptera heteroptera.

In examining sections of female larvæ stained with safranin-gentian-violet, I was surprised to see a very marked polarized or bouquet stage and to find among the loops something resembling the odd chromosome of the growing spermatocytes. It was difficult to get a clear view of this body as it lay within the loops. In one section of a slightly earlier stage before synapsis, there were found two pairs of chromosomes (fig. 271, x₁, x₂, and m₁, m₂) which were stained with safranin in contrast with the violet spireme. These two pairs I interpret as being (1) the homologues of the pair of m-chromosomes, which remain condensed during the growth stage of the spermatocytes, and (2) a pair of heterochromosomes corresponding to the odd chromosome of the male. Various combinations of these heterochromosomes are shown in figures 272-277. Figures 278 and 279 were taken from mercuro-nitric material stained with iron-hæmatoxylin. In section 278 the "bouquet" was cut through, showing the bivalent corresponding to the larger pair in figure 271, and in figure 279 this element is seen behind the paler loops. The history of these two pairs of heterochromosomes, which have not, so far as I know, been found before in oöcytes, should be followed up in older ovaries, and related species should be examined for similar phenomena.

LEPIDOPTERA.

Cacoecia and Euvanessa.

I had no intention of making an extended study of the spermatogenesis of the Lepidoptera, but was interested to see if anything corresponding to the heterochromosomes of other orders could be found. The material studied was the testes of the larvæ of Cacœcia cerasivorana and Euvanessa antiopa. The number of chromosomes is large, but the equatorial plates are diagrammatically clear. In both species 30 chromosomes are found in both first and second spermatocytes. In both, one chromosome is larger (figs. 290 and 293, x). In the growth stage (figs. 283, 284) there is a two-lobed body (or sometimes two separate spherical bodies) which seems to correspond in size to the larger pair of chromosomes in the first spermatocyte. In iron-hæmatoxylin preparations this pair is often obscured by parts of the spireme which are tangled around it. In safranin-gentian preparations it stains, not like a plasmosome, but red like the heterochromosomes, while the spireme is violet. The staining reaction at least suggests that this equal pair of chromosomes, which may be traced through the synizesis stage (fig. 280), synapsis stage (figs. 281, 282), growth stages (figs. 283, 284), and prophases (figs. 285-287), into the first spermatocyte spindle (figs. 288, 290), and on to the second spermatocyte (figs. 289, 291, 292), is an equal pair of heterochromosomes comparable to the equal pair of "idiochromosomes" found by Wilson in Nezara ('05). As the various stages are practically the same in Euvanessa antiopa, but somewhat clearer in Cacœcia, only one figure is given for Euvanessa—the equatorial plate of the first spermatocyte (fig. 293).

SUMMARY OF RESULTS.

(1) An unequal pair of heterochromosomes has been found by the author in 19 species of Coleoptera belonging to 8 families: