26 Papers from the Marine Biological Laboratory at Tortugas. 
a giant spermatid (with the accessory chromosome, of double size in this 
case) in process of metamorphosis, and a fully formed giant spermatozoon. 
Giant spermatozoa have been frequently observed among the insects. 
Wilcox (1895) found them extensively and made a study of them in 
Cicada tibicen and Caloptenus femur-rubum. With Wilcox, I believe that 
they are non-functional also in Aplopus mayeri, and that “they are excluded 
from the developmental series and really come to nought.” 
I have noted above that the primary spermatogonial cell frequently 
divides amitotically. This may occur several times, giving rise to a multi- 
nucleate cell. Contrary to what Wilcox has found in Cicada tibicen, where 
the giant spermatozoon “arises directly from spermatogonia without cell- 
division, by a metamorphosis of the nucleus,’ figure 24 shows that the 
binucleate cell resulting from an amitotic nuclear division may subsequently 
divide karyokinetically. Such a cell would give rise to spermatocytes of 36 
chromosomes (which have been observed) and eventually to giant sperm- 
atozoa. I have not observed spermatocytes with 72 chromosomes, but such 
may very well arise as a result of two successive amitotic nuclear divisions. 
Frequently spermatids are seen with two or even several tails. This 
phenomenon is due probably to an accidental or abnormal division of the 
centrosome, from each product of which an axial filament grows out. Adult 
spermatozoa thus deformed are only seldom seen; they probably early 
undergo degeneration. 
THEORETICAL CONSIDERATIONS. 
INDIVIDUALITY OF CHROMOSOMES. 
Among the ordinary chromosomes morphological individuality can not 
be convincingly demonstrated. This is due to the fact that between every 
mitosis, both spermatogonial and spermatocytic, as well as previous and sub- 
sequent to synapsis, a brief resting stage is interpolated when the chromo- 
somes are merged into the nuclear reticulum. Correspondence of size can 
readily be found between the chromosomes of the equatorial plates of the 
primary and secondary spermatocytes, as also between these and pairs of 
chromosomes of the spermatogonial mitoses, but I do not consider the corre- 
spondence sufficiently close or striking to contribute reliable evidence in 
favor of the above hypothesis; nor do I believe it possible to find very strong 
evidence from this source in cases where we are dealing with so large a 
number of chromosomes. What evidence there is, however, points in the 
proper direction, as will be noticed by comparing figures 17, 74, and 114. 
The evidence yielded by the accessory chromosome, however, is definitely 
corroborative of this hypothesis. When once fully differentiated in the later 
orders of the secondary spermatogonia it retains thereafter a persistently 
definite shape, size, and location in the nucleus, and never passes into a 
reticular stage. Even when assembled among the ordinary chromosomes of 
