914 
cessiveness represent the extremes of pre- 
potency and subordination in the relations 
of the diploid germ plasm. 
Such evidence as this suggests that the 
germ plasms are related to each other 
much as competing races of organisms are. 
The evidence here may be said to favor the 
idea of the germ plasm as a mass of inde- 
pendent individual corpuscles competing 
with each other for their existence. One 
of the most obvious ways in which Men- 
delian concepts have influenced the idea of 
the germ plasm is in the emphasis which 
the idea of Mendelian dominance gave to 
the conception of antagonistic relations 
between its elements. Of any two char- 
acters present in the parents, one may 
dominate the other in the offspring with- 
out destroying it. These facts are ob- 
viously opposed to a chemical theory of the 
germ plasm. If the union of the two fac- 
tors is chemical it should result in some 
new compound with properties in some 
degree different from either. As has been 
pointed out by Friedmann on chemical 
analogy, the product of the union of two 
germ plasms would not necessarily be in- 
termediate in its properties between those 
of the two combining elements, especially 
in its form and spatial configuration. The 
angles of a crystal made of a mixture of 
two isomorphic salts are not the average of 
the angles of the two components. Such 
so-called chemical theories of dominance as 
assume the existence of regulative enzymes 
influencing the rate of morphogenetic 
processes by their relative concentration 
seem to me to beg the whole question by 
assuming in the appearance of the regula- 
tive enzymes at the critical point the ele- 
ment of organization which they start out 
to deny. 
Not only the facts as to constaney and 
variation in the number of chromosomes, 
but also our knowledge of their constant 
SCIENCE 
[N.S. Vou. XXXV. No. 911 
arrangement and position in the cell nu- 
cleus, has been greatly strengthened by 
the critical study of recent years. The ac- 
cumulating evidence for the constant posi- 
tion of each chromosome in the resting 
nucleus as it reappears after the disturbance 
of nuclear division is most suggestive of 
mechanical organization in the nucleus as 
a whole, whether or not there be an ultra- 
microscopic organization of the chromo- 
somes themselves. 
Boveri has shown for Ascaris that the 
long chromosomes of the embryonic nuclei 
reappear at each cell generation in the 
same position and determine the configura- 
tion of the young resting nuclei. In cer- 
tain fungi at least there is permanent 
connection between the chromosomes and 
centrosomes, thus determining the polar 
organization of the cell as conceived by 
Rabl. These conditions give strong evi- 
dence for a mechanically organized, rather 
than a mere chemical cell and nuclear 
structure. 
No more important question confronts 
the students of the cell to-day than the de- 
termination of the relative positions of the 
paternal and maternal chromosomes in the 
diploid vegetative cells. Evidence that the 
serial arrangement found in all spirems is 
maintained in the resting condition and in 
the equatorial plate of a species of Carex 
will be presented at this meeting. Stras- 
burger, Overton and others find that the 
chromosomes are paired throughout veg- 
etative development, the members of the 
pairs representing respectively maternal 
and paternal elements. If these results 
are confirmed and found to be general, we 
have a further vantage point from which 
to attack the question as to the method by 
which the parental germ plasms are not 
only transmitted but come to expression 
in the offspring. 
It is fairly well established that homol- 
