78 BULLETIN: MUSEUM OF COMPARATIVE ZOOLOGY. 



sists entirely in its further elongation and in the continued develop- 

 ment of the spiral. Thus in the mature spermatozoon (Figs. 120, 

 121) the spiral is so pronounced that a straight line drawn through the 

 centre of the head would barely fall within the limits of this structure 

 throughout its entire course. The unstained central portion of the head 

 is present, just as in preceding stages, and may be observed buth in cross 

 and longitudinal sections, and is occasionally shown even in surface views 

 (Fig. 120). 



In these later stages of metamorphosis the centrosomes are still readily 

 to be demonstrated. The two lateral ones lie in the position that charac- 

 terized them at previous stages, i. e. side by side at the base of the head 

 and enclosed in the nuclear membrane of the head. Extending back- 

 ward from these two centrosomes is a triangular structure, evidently 

 made up of fine fibres, which joins the lateral centrosomes with the end 

 of the axial filament (Figs. 116-120). The triangular region thus en- 

 closed stains much less deeply than the axial filament proper. In some 

 cases there seem to be two very distinct fibres at the border of the tri- 

 angle (Fig. 119), but such appearances are rare. This triangular region 

 varies in its proportions, being usually long, with the acute angle directed 

 toward the axial filament (Figs. 117, 118, 120), while occasionally it 

 may be more nearly equilateral (Figs. 116, 119). When the spermato- 

 zoon is viewed parallel to the plane of the three centrosomes (Fig. 120), 

 no evidence of this triangular region is to be observed, showing beyond 

 a doubt that the triangle represents a membrane or more probably a 

 number of fibres lying in one plane and spreading out from the base of 

 the axial filament like the ribs of a fan, to join the lateral centrosomes 

 at the base of the head. 



It has been noted that at earlier stages there is always an enlarge- 

 ment at the basal end of the axial filament. This structure, which is de- 

 rived directly from the original spermatid centrosome, retains its close 

 connection with the axial filament throughout all the transformation 

 stages, and may often be seen in the fully developed spermatozoon as a 

 very slight enlargement at the apex of the triangle described above (Fig. 

 116). In many spermatozoa, however, this element cannot be observed, 

 and it is always very inconspicuous. From its long-continued persist- 

 ence throughout all the transformations, and its presence in many of the 

 mature spermatozoa, I believe there can be no doubt of its importance. 

 In position and origin it corresponds to the end knob in the spermatozoa 

 of many animals. 



One of the most remarkable facts concerning the mature spermatozoa 



