538 Lewis . — The Behaviour of the Chromosomes 
in the spirem and that they may either give rise to rings by looping on 
themselves (Fig. 31) or dumb-bells by cross-segmentation without looping 
(Fig. 28) ; the cardinal points of this being, that no conjugation of spirems 
takes place before synapsis, that the chromosomes are united end to end to 
fcrm the bivalents, and no longitudinal split is observed. A similar method 
of conjugation of chromosomes has been given by Strasburger (’04) for 
Galto7iia and by Mottier (’07) for Tradescantia. 
The chromosomes are now located mainly near the nuclear membrane 
which has up to this time retained its characteristic appearance. The 
chromosomes are connected more or less by thin delicate strands which 
often reach out to the nuclear membrane or even to the cytoplasm. The 
nuclear membrane now begins to fade out and spindle fibres crowd into the 
cavity. At this time the segments have shortened almost to their final 
form, although there is yet considerable difference in their relative sizes. 
The segments in Tfaja become relatively much shorter and thicker than 
in Pinus , and accordingly the stages following spindle formation are much 
more difficult to follow than in Pinus. The orientation of the chromosomes 
into the equatorial plate now takes place. This occurs quite rapidly and 
the chromosomes become arranged almost in a single plane. This is quite 
strikingly true of Thuja. The structure and relation of segments is not so 
apparent at this time, but a careful examination reve'als all of the forms 
observed in the prophase. The chromosomes are closely crowded together 
and some are in a plane slightly above or below the rest. There is no 
indication that the segments have begun longitudinal fission as yet (Fig. 20). 
The stage of the spindle plate lasts some time in TJnja , as the division goes 
on comparatively slowly in this form. The spore-mother-cells undergo this 
division about the first week in February. Entire sporangia have been 
observed in which every cell exhibited the chromosomes in the plate of the 
spindle. Such cones brought into the laboratory and placed in a warm place 
finish the division in a few hours. 
The manner of attachment of the spindle fibres, as well as the behaviour 
of the chromosomes during the metaphase and anaphase, was found to be in 
general harmony with the well-known accounts of Strasburger (’00) and 
Mottier (’03). Miss Ferguson (’04) states that the spindle fibres are never 
attached at the free ends of the segments, but invariably at some point along 
the fused part, but this does not appear to me to be the case. The chromo- 
somes are usually arranged tangential to the spindle and sheaves of fibres 
are attached to the segments either at the free ends or at the point of 
adherence of the two segments. A few of the chromosomes stand radially 
to the spindle. 
The separation of the daughter segments now takes place. Two well- 
known theories are still held as to the manner in which the separation is 
effected. Miss Ferguson (’04) maintains that the sister segments are not 
