284 
R. A. HARPER 
Further, they reproduce only by the division of parent chromosomes — a 
division which again outside the cases of certain protozoans gives evidence 
of its painstakingly equational character. We have then in the chromo- 
somes subsidiary bodies in the protoplasm which duplicate in their perma- 
nence and method of reproduction the cell itself. As Boveri has empha- 
sized, each chromosome has a life history of its own. It divides for repro- 
duction, and each daughter chromosome grows to the size of the parent 
chromosome and in cyclic fashion is again divided. This is the basis for 
the doctrine of the individuality of the chromosome and has been used in 
support of the much vaguer, panmeristic doctrines of cell division and of 
the older theories that the cell is a complex of lesser life units, each, however, 
endowed with the full complement of the essential cell characters — the 
capacity to assimilate, grow, and divide cyclically into equivalent daughter 
units. These latter conclusions, however, it seems to me far outrun the 
facts. The data from elaioplasts, plastids, vacuoles, nucleoles, etc., suggest 
that, while permanence and reproduction by division can be present in 
various degrees in al' these structures, they are rather acquired conditions 
worked out and developed in connection with the metabolism of the cell 
as a whole than essential characteristics of life units which by their com- 
bination then make the assimilation, growth, and division of the cell as a 
whole possible. 
The permanence and equational division of the chromosomes are none 
the less the great outstanding discoveries of cytology as to cell organization, 
and here, as in the case of the constriction of the plastids, the underlying 
physical and chemical phenomena involved have so far escaped discovery. 
Neither chemistry nor physics furnish any data which aid the cytologist 
to discover why the segments of the spireme thread should split longitudi- 
nally or why the spireme should divide transversely to form the chromo- 
somes. For the separation of the daughter chromosomes we have the 
possibility of contractility in the spindle fibers, protoplasmic streaming 
even, etc. — processes with adequate physical analogies — but the initial 
splitting of the chromosomes before the spindle is formed is quite without 
parallel in the behavior of atoms, molecules, or larger colloidal particles 
as known to the chemist and physicist. The division of plastids, chromo- 
somes, and, as we shall note presently, centrosomes, as well as the pairing 
of chromosomes in cell and nuclear fusion, are basic data of cytology that 
so far differentiate cellular organization and processes from those of unor- 
ganized matter. 
As to the functions of the chromosomes, we have a mass of evidence, 
which has been so many times and so well summarized that I need not 
rehearse it here, that they in some way provide for the so-called transmission 
of the hereditary characters. The evidence that the chromosomes are a 
physical basis for heredity is, it seems to me, entirely convincing and ade- 
quate. It is particularly significant that with the intensive study of heredity 
