Pollen Mother-cells of Lilinm canadense . 203 
nuclei, and from a study of such sections it is quite certain that the thread 
at this stage is continuous. 
I have been able to find no regularity whatever in the course of 
the thread, except that it seems to be as evenly distributed as possible 
throughout the nuclear cavity. Sections sometimes give the impression of 
a greater massing toward the centre of the nucleus, but careful study shows 
this effect to be due to the greater number of thicknesses of the thread to 
be seen in looking through the middle part of the spherical nucleus. The 
portions of the thread which reach the periphery and then run back into the 
interior of the cavity might be described as loops, but in the arrangement 
and course of these loops there seems to be no regularity. There is no 
apparent relation between the number of loops at this stage and the number 
of chromosomes which are to be formed later. There are certainly many 
more than twelve peripheral portions of the thread, and in some cases 
I have been able to count more than twenty-four, although the exact 
determination of the number in any nucleus is extremely difficult ; but 
as yet there seems to be no relationship between any definable portions of 
the thread and the individual segments into which it is later to be divided. 
Neither at this nor at any later stage of this division have I been able 
to detect any evidence, in the arrangement of the spirem or of the chromo- 
somes, of a polarity of the nucleus, such as was described by Rabl (’85) for 
the prophases of nuclear division in animal cells, and as was found by 
Flemming (’87) in the spermatocytes of Salamandra. In this respect the 
pollen mother-cell of the lily offers a striking contrast to its immediate 
ancestors ; in these, the spirem, before segmentation, has a very uniform 
arrangement, corresponding closely with that of the daughter-spirem of the 
previous division. Strasburger (’04), however, finds that in the pollen 
mother-cell of Tradescantia the spirem is arranged in a very regular spiral ; 
and both Strasburger (’88) and E. Overton (’91) have described an arrange- 
ment of the chromosomes and the nucleole in pollen mother-cells with 
reference to a ‘ Polfeld * and a ‘ Gegenpolseite.’ 
The external projections of the chromomeres show best in thinner 
sections stained with the triple stain, such as the one represented in Fig. 22 . 
It is quite conceivable that the fine fibres, which run out for a short distance 
from these projections, may belong to a system which connects all parts of 
the thread with one another, and possibly with the cytoplasm as well ; 
in such a case the greater part of the fibrous connexions may be either too 
delicate to be differentiated and distinguished with our present technique, or 
may have been destroyed in some of the processes involved in making the 
preparations. I have seen no direct evidence of such connexions in the lily ; 
but similar systems of fibrous connexions between the chromomeres of 
various parts of the spirem have been found by Brauer (’93) in the sperma- 
togonia and the primary spermatocytes of Ascaris , by Winiwarter (’00) in 
Q 
