214 
Allen . — Nuclear Division in the 
have undergone as they are being pulled apart. In some cases (Fig. 64) 
there is an appearance which may be interpreted as that of four ends 
turned toward the observer, especially when viewed under medium magnifi- 
cation. But careful study of such a figure with a higher power shows that 
two of the four parts are at a higher plane and are really the ends of the 
daughter chromosomes, while the other two are in a lower plane and 
represent simply lateral protrusions of the daughter chromosomes due 
to their twisting about each other. Fig. 64 shows how such a four-parted 
appearance may arise. As already remarked, I have never found an 
instance of an actually four-parted chromosome (i. e. one showing the 
second longitudinal split) in the equatorial plate. 
In case the radially arranged chromosomes are seen laterally (Figs. 
60-63), their two-parted nature is even more plainly evident. In this 
position we can note the variations in form already described ; the daughter 
chromosomes may be somewhat twisted about each other, with their outer 
ends either diverging (Fig. 60), in contact (Fig. 63), or apparently fused 
(Fig. 62) ; or the daughter chromosomes may be simply in contact, with 
little or no twisting (Fig. 61). 
As has been said, the majority of the chromosomes are of the type 
already described, namely radially arranged and attached by their inner 
ends to the periphery of the spindle. A few, however, which lie peri- 
pherally are tangential to the spindle. These tangentially arranged 
chromosomes are always, so far as I have observed, of the twisted, closely 
appressed type. The attachment of the spindle fibres to the daughter 
chromosomes in such a case is not necessarily at the end ; it may be at 
either end, or at or near the middle, or at any point between the middle 
and either end. Figs. 65-67 furnish a series, showing this variety in the 
point of attachment of the spindle fibres. The attachment to the daughter 
segments of any parent chromosome is always, however, at corresponding 
points. 
The chromosomes lying in the interior of the equatorial plate ( b , c , d , 
Fig- 57), like the tangential chromosomes, typically consist of twisted, 
closely appressed segments ; and, again like the tangential chromosomes, 
the attachment here is not at any definite point, but may be anywhere 
along the length of the segment. 
To sum up the condition of the chromosomes as regards their attach- 
ment to the spindle : in a large majority of cases the fibres are attached 
at, or very near, one end of each daughter chromosome ; but they may be 
attached at the middle, or at any point between the middle and either end. 
It will be shown that the differences between the forms exhibited by the 
daughter chromosomes in the metaphases and anaphases result from this 
variation in the point of attachment of the spindle fibres. 
