ON SPERMATOGENESIS IN THE RAT. 
361 
vertebrate supporting cell, has a different morphological sig- 
nificance, the supporting cells being probably derived from 
follicular cells which appear not to be represented in the in- 
vertebrate testis. 
As soon as the separation of the globules has taken place, 
or even before this, the spermatozoa begin to travel bodily 
towards the lumen of the tube. This movement is apparently 
produced by the supporting cells, which convey the sperma- 
tozoa inwards to the lumen of the tube, where they finally 
become detached. A supporting cell which is thus casting off 
its group of spermatozoa, is represented by fig. 21. For a 
short time the head of the spermatozoon remains embedded in 
a protoplasmic envelope, perhaps derived from the supporting 
cell, and a small granule of protoplasmic material, darkly 
stained in gold preparations, remains for some time at the 
junction of the head with the middle piece (fig. 24, b, c), but 
eventually disappears. 
A mature spermatozoon, examined fresh, or mounted in gly- 
cerine after osmic acid, shows no trace of a division into body 
and tail, appearing to be composed of two parts only, the head 
and the long tapering body ; but by treatment with chloride 
of gold, as before described, the division into middle piece and 
tail is rendered very conspicuous. 
Fig. 25 represents a spermatozoon from the epidymis, which 
is mounted in glycerine after having been treated with chlo- 
ride of gold. The middle piece is somewhat swollen and 
stained by the reduction of the gold, the staining being chiefly 
concentrated in a fine spiral fibre, which winds closely round 
this portion. It has a length of about ’07 mm., and presents a 
striking contrast to the tail, which is absolutely unstained ; 
the length of the tail is about -08 mm. The spiral filament, 
as seen in sections mounted in balsam, has been already de- 
scribed (fig. 12, d). I have not been able to make out the 
manner of its development. It is first seen when the sperma- 
tozoa have reached their full length, before they have begun to 
travel to the lumen of the tubule (fig. 14) . 
