Information Content and Biotopology of the Cell in Terms of Cell Organelles 223 



decisions are represented as branch points, which may or may not be taken ; thus 

 when the nebenkcrn (.1001) develops from the fusion of mitochondria (.100), 

 the mitochondrion loses potential on making this decision, but suffers no loss in 

 faihng to do so. 



Admittedly several apparent and some real inconsistencies exist in this 

 provisional flow-diagram. For example, the development of the cilium from a 

 mitochondrion-hke body has already been suggested (34) in which the cilium 

 should appear related to, but more specified than, its supposed progenitor; the 

 present tree only remotely relates these. This difficulty may be resolved in a 

 common progenitor organelle (.1) at the plasmic-interfacial decision point. 

 Another inconsistency is in the separation at 0. 1 11 into cirrus-peniculus-hexagon 

 complexes vs. ribbed wall and brush border, the former being ciliated, and the 

 latter non-ciliated organelle-complexes. A more reaUstic decision in Paramecium 

 might be to form pellicle system unit (hexagon-rhombus complex) or to form 

 gullet system unit (peniculus-quadrulus-ribbed wall) dependent perhaps on loca- 

 tion of the primary organelle at a region of pattern homogeneity (amongst 

 hexagons or amongst rhomboidal elements) or at a region of pattern contrast (at 

 the junction of hexagons and rhomboids {kj to kj — 1 in Fig. 3, discussed in the 

 next section)). The latter condition is consistent with Paramecium structure, but 

 so far has been demonstrated only in the ciliate Stentor (35). 



Several investigators (25, 36, 37, 38) have related nuclear membrane and 

 endoplasmic reticulum as mutual derivatives — a view not inconsistent with the 

 present scheme in its broadest sense; in addition to the rounds of packing and 

 fusion, some 'unravelling' may be involved in both cases. Viruses are pro- 

 visionally included in the tree because they nearly satisfy our definition, because 

 of their organelle-like ultrastructure, because of recent evidence for host related- 

 ness (40, 41, 42) and because of frequent nucleolar involvement (43, 44, 45). If 

 considered as particles produced by the host's gene-product synthesis that 

 contain a small error perpetuated by error feed-back, then the nuclear viruses 

 might occupy the packed (46) small-nucleolus position (.001); the cytoplasmic 

 viruses would be classed with 'assorted granules' (.1000) if single, and along 

 with 'grana' and 'fibers' (.101) if packed. It is not necessary to postulate a 

 'nuclear round' for the replication of either extranuclear organelles or of 

 cytoplasmic viruses, although this may be the usual case. 



IV. PRIMORDIAL GRAPH 



A material basis for nucleocytoplasmic communication is therefore realized 

 in the cell organelle, which in its primitive unspecified state provides the cell 

 with its necessary potential of structural diversity through relatively few bits 

 of information. A further quality related to the above (47) is that the cell's 

 geometry of functional relations is in some respects as similar to its cytological 

 structure as a wiring diagram is to its final construct. In other words, at the 

 level of the cell organelle (above the macromolecular and below gross cellular 

 dimensions), a coincidence between the functional and the structural biotopo- 

 logical set points of the cell exists. These relationships are represented for 

 organelles at an intranuclear-extranuclear decision in Fig. 3a, and for functioning 

 cytoplasmic organelle systems (48) in Fig. 3b. The intranuclear-extranuclear 



