4o6 
AMERICAN JOURNAL OF BOTANY 
[Vol. 9. 
densed, the chromosomes found themselves Hnked in this very definite 
way. It is not Hkely that a chromosome could pass between two members 
of a bivalent pair in the way seen in this plant, if those two mem- 
bers originally lay side by side in the same spireme. The univalent chro- 
mosomes making up a bivalent, if attached at all during the open spireme, 
must have been attached end to end. 
On the whole, therefore, the singular absence of parallelism at all stages, 
the clear relationship established between the second contraction loops 
and the bivalent chromosomes, and the regular linking of the latter during 
diakinesis, seem to make it quite certain that we are dealing here with 
chromosomes which are attached end to end in a univalent spireme. 
The Position of the Chromosomes. The regularity with which the biva- 
lent chromosomes are linked together in diakinesis raises an interestmg 
question concerning the relative positions of the chromosomes in the nucleus. 
The fact that linking seems to conform to a definite scheme suggests that 
during the preceding spireme stage the individual chromosomes occupied 
the same relative position in all nuclei, being attached to each other at 
the same places and in the same way. How far back this condition ex- 
isted is probably difficult, if not impossible, to determine in a plant where 
the individuality of the chromosomes is lost sight of during resting stages; 
but at least the interesting possibility is suggested that such a definite 
arrangement might be present in all nuclei. Whatever may be the situation 
in the somatic cells in general, it is quite probable that the chromosomes 
are arranged during the last archesporial mitosis according to a definite 
plan, so that they come to occupy the same relative positions in the sub- 
sequent prophase stages. 
The Nucleolus. There is always at least one nucleolus in the nucleus 
of every pollen mother cell in Oe. franciscana. That its function seems 
to be at least partially that of storage is indicated by the appearance pre- 
sented during late synizesis, when the threads surrounding and attached 
to the nucleolus become greatly swollen, as though gorged with material 
suddenly poured into them. At the same time, the nucleolus begins to 
lose its staining qualities and to look empty. The material which it stores 
stains intensely with the heamatoxylin, and would seem, therefore, to be 
a form of chromatin. The nucleolus itself, when divested of the chromatic 
contents, does not take the stain and would therefore seem to be composed 
of a substance perhaps closely related to, or identical with, linin. At 
different times it presents a somewhat different appearance. Before dis- 
charging its contents, it contains at times a crystalloid body. In other 
cases one or more vacuoles are present in the interior. After it has been 
emptied of its chromatin content, it very generally shows within a black- 
staining, spherical little body, the endonucleolus. This seems at times to 
be connected directly with the spireme by a prominent thread, but usually 
it lies freely in the nucleolus. It outlasts the nucleolus ordinarily, being 
