Balls .— The Mechanism of Nuclear Division. 655 
during its phylogeny it has probably lost all other powers and become 
a passive organ, handed about by other organs which are the blind agents 
for the distribution of the controlling substance, although they themselves 
are incapable of directing the fortunes of the cell. 
Some such unprejudiced mechanism is required to explain the Men- 
delian facts of gametic distribution, and the achromatic structures seem to 
be this mechanism. 
The only satisfactory explanation seems to be given by admitting the 
reality of the spindle fibres, and by assuming the power of contractility in 
the achromatic structures generally. There is no assumption of vitalism in 
the latter step, for the fact of fibrillar contractility is fully acknowledged 
in pseudopodia and cilia. There is, moreover, a probability that future 
physical research upon surface tension in thin films and threads may be 
able to give us a physical explanation of the phenomenon. Contractility 
in the spindle fibres was first advocated by Klein and Van Beneden ; Boveri 
demonstrated that the fibres became shorter and thicker ; Hermann showed 
that the central spindle fibres elongated in mitosis. Hertwig and Wilson 
were unable to see any increase in the thickness of the fibres, but it must be 
remembered that these fibres are so delicate that one may be unable to 
distinguish the diffraction image of a thin one from the real image of 
a thicker one. 
Strasburger has cleared the way for a conception of motor mechanism 
in cell-division, or ‘ kinoplasm but he does not differentiate between the 
nuclear and cytoplasmic portions, whereas I have reason to consider the 
two forms as morphologically distinct, though perhaps chemically similar 
or identical. 
The Centrosome . 
A persistent nucleolo-centrosome which divides to form a spindle is 
found in some unicellular organisms. This body is considered as equivalent 
to the centrosome. The * end-plates ’ of Actinosphaerinm arise by the 
division of a nucleolus or plasmosome, and it would seem from the general 
evidence of these primitive organisms that the achromatic plasmosome and 
the centrosome have a genetic relationship. This is even more distinct in 
the case of Par amoeba, whose ‘ Nebenkorper ’ consist of extruded achro¬ 
matic substance. From this we pass to Hertwig’s work on Actinosphaerinm , 
where the centrosomes are proved to arise from the reticulum which is 
extruded at the poles, and this agrees with the increasing number of cases 
recorded in which the centrosomes are formed anew. 
When asters are found in organic cells they generally seem to be of 
cytoplasmic origin, and independent of the centrosome. The observations 
recorded below have led me to the opinion that the asters are heterogeneous, 
the outer portion being merely stream-lines in the cytoplasm, while the 
remainder of the rays are true fibres, derived from the achromatic nuclear 
