WILCOX: SPERMATOGENESIS OF CALOPTENUS FEMUR-RUBRUM. 197 
elements of each resultant spermatid present the appearance of a com- 
pact irregularly shaped nuclear mass (Plate 2, Figs. 41, 53). After the 
formation of a nuclear membrane there is a considerable swelling of the 
nucleus as a whole, and the chromosomes break up into a large and 
variable number of granules of unequal size (Plate 3, Figs. 80-86 and 
103, 104). The fine granules later fuse into a homogeneous crescentic 
mass (Fig. 105). Cross sections of the head of the spermatid appear 
circular (Fig. 110*). There is, therefore, no flattening of the head in 
Caloptenus. 
The above results, thus briefly stated, may now be compared with the 
investigations of other students published since my former paper was 
written. 
Rickert (94) has found in the odgenesis of Copepods many stages 
in the origin of the Vierergruppen similar to what I find in the sper- 
matogenesis of Caloptenus. (Compare my Figs. 9 and 15-17 with his 
Figs. 23, 24.) He finds that there is one longitudinal splitting of the 
chromatic segments. Then each of the daughter segments divides 
transversely. On this point Rickert says (pp. 308, 309): “Sie [the 
transverse division] tritt, wie erwahnt, im Stadium der Fig. 11, also 
dann, wenn die Chromosomen gegen den Aequator des Keimblaschens 
vorzuriicken anfangen, deutlich hervor als eine allen Doppelstiiben ge- 
meinsame Erscheinung. Indessen lisst sich die Vorbereitung zur Quer- 
spaltung schon in jiingeren Stadien (Figs. 8-10) erkennen, an manchen 
Doppelstaben sogar recht deutlich.” Since I do not find a longitudinal 
splitting, it may seem difficult to compare Riickert’s account with mine. 
In reality, however, a very close agreement exists. Riickert states that 
ordinarily the transverse division is first manifest when the pairs of 
segments begin to move toward the equator of the spindle. This corre- 
sponds to my second mode of formation of Vierergruppen as described 
and figured (Diagrams 7-10) in my former paper (Wilcox ’95, p. 10). 
In this case the four ultimate chromatic elements of a ring are not to 
be distinguished until the rings are about to take their place on the 
spindle. While Riickert admits that his transverse division is indicated 
much earlier in some cases, he adds (pp. 308, 309): “ Vollends aber 
wiirde es falsch sein, in den Fiden noch jiingeren Stadien, in welchen 
sich das Chromatin noch nicht konzentriert hat, alle Unterbrechungen 
des firbbaren Teiles als Vorlaiufer der spateren Querteilung zu betrachten. 
Eher liesse sich die in friiherer Zeit, namentlich in Fig. 8, so deutliche 
Ansammlung des Chromatins in den anschwellenden Fadenenden viel- 
leicht mit der spiiteren Querteilung in genetischen Zusammenhang 
