1905.] NATURAL SCIENCES OF PHILADELPHIA. 173 



corroborative of my (1900) earlier ones upon Peripatus. In the 

 achromatic spindle a central spindle, fibres continued from pole to pole, 

 but not attached to chromosomes, is found in the spermatogonia, but 

 not in the spermatocytic mitoses; its fibres and the pole fibres are 

 formed from the cytoplasm. The mantle fibres, on the other hand, 

 those connecting the chromosomes with the centrosomes, are derived 

 from linin fibres previously connected with the chromosomes — the 

 mantle fibres are at least in great part nuclear in origin. Whatever 

 be our views upon the nature of the pole and central fibres, whether we 

 regard them as lines of currents or as actual fibrils, I think we must 

 consider the mantle fibres as contractile fibrils, not simply paths of 

 movements of fluids. This follows clearly from the results of my Per-i- 

 patus paper, where the mantle fibres of mitosis were shown to be 

 derivable from linin fibrils stretched out through the nucleus in the 

 rest stage, and there constituting a continuous linin thread (spirem) 

 with many fine collateral branches. Such fibrils crossing one another 

 in all directions in the resting nucleus cannot be considered current- 

 paths; how, then, change into current-paths during mitosis, except in 

 so far as we regard a contracting gum-elastic cord to be a path of move- 

 ment? What holds for the mantle fibres need not, however, obtain for 

 the pole and central spindle fibres. The other achromatic spindle 

 constituents of nuclear origin are the connective fibres, fibres pulled 

 out between two separating daughter chromosomes; these are clearly 

 derived from the linin forming the matrix within which the chromatin 

 of a chromosome is imbedded or the sheath by which it is surrounded. 

 When two daughter chromosomes separate in metakinesis it has the 

 appearance as though two connective fibres pass between them; but it 

 is more likely that such two lines represent in actuality the boundaries 

 (visible because of their higher refraction) of a solid or hollow linin 

 cylinder. 



Already in the monaster stage of the spermatogonia (Plate IX, fig. 9) 

 as of the first spermatocytes (fig. 32) there is a pair of centrosomes at 

 each spindle pole . After each spermatogonia division these centrosomes 

 wander through an arc of 180° to take up a position on the opposite 

 side of the nucleus ; the same process seems to take place in the sper- 

 matid ; but before the second maturation mitosis each wanders through 

 an arc of only 90°. In the first maturation mitosis there are two man- 

 tle fibres from each spindle pole to each chromosome (Plate IX, figs. 

 32-33; Plate X, figs. 34-36); in the second only one (fig. 39). This is 

 understandable on the basis that the first mitosis is reductional, 

 since, as is most clearly shown in a straight dumbbell-shaped bivalent 



