A ORITICISM OF THE CELL-THEORY. 161 
das der als Karyokinese bezeichnete Vorgang nicht ein spe- 
cielle Kerntheilungsmodus, sondern der Kerntheilungsmodus 
kar’ éEoynv ist”? 
I think not. Particles of chromatin scattered through the 
protoplasm do not constitute a nucleus any more than a heap 
of bricks constitutes a house. Under such a view, Ciliata like 
Trachelocerca phenicopterus and Chenia teres would 
not be cells, for they have no central nucleus of complex struc- 
ture, nor have Oscillaria and Bacterium, in which chromatin 
granules have been discovered. Though the case of Holos- 
ticha scutellum, in which scattered nuclei (chromatin par- 
ticles) unite and fuse to form a single central body or nucleus 
previous to division, may help to clear our ideas, it is evident 
that the demand for a central organised constituent is 
more than the cell conception can bear, especially if the 
demand carries with it a further demand for the universality 
of mitotic division in nuclei. 
In short, before we could accept Hertwig’s definition of 
a cell, we should have to ask and answer the question, What 
is a nucleus? 
Here I may stop to ask whether it is worth while to discuss 
the grounds of a definition which, when made, could not be 
acceptable to the mind of everyone. An argument about 
definitions would soon land one in the regions of scholasticism, 
and I have no desire to enter into subtleties which would tax 
the powers of a Duns Scotus. To give an answer which shall 
be beyond all cavil to the question, What is a nucleus? would 
be about as easy as to answer how many angels can dance on 
the point of a needle. 
The truth is that it is the attempt to frame short concise 
definitions, applicable without exception to whole classes of 
phenomena, which leads to trouble. The concepts of biology 
may and should correspond with the phenomena we observe, 
but they can very seldom be made into universal propositions. 
There is no place in the science for definitions as exact and 
universal as those of geometry. The qualities of a nucleus are 
not to be defined like those of a point or a line. Such propo- 
vou. 388, PART 1,—NEW SERIES. L 
