537 
in Pimts and Thuja . 
nature is concealed. From a study of the chromosomes just after cross- 
segmentation of the spirem and before the segments have thickened to their 
final form it becomes evident that two pieces have become in some way 
approximated to form the bivalents. That these two members are the 
sides of the loops there seems little doubt. There are, however, many 
nuclei in which such a looping of the spirem with such diagrammatical regu- 
larity cannot be observed. In such nuclei the looping is far less regular. The 
spirem often has not become entirely regular, but has retained to a con- 
siderable extent its lumpy nature and the longitudinal split has not entirely 
disappeared (Fig. 13). The spirem consists of short loops, but there are 
often stretches of considerable length in which there seems to be no ten- 
dency to form loops. These pieces often reach almost across the nuclear 
cavity (Figs. 13, 16). That the chromatin content is being drawn out of the 
parts of the cord and deposited in other is often quite evident (Fig. 15). 
This often presents the appearance of dark, deeply staining parts connected 
by thin strands, the whole appearing like a reticulum (Fig. 15). It seems 
that the spirem breaks up and pieces are brought together and fuse. There 
is no such massing of the spirem in this condition to a central aggrega- 
tion as is often apparent in other nuclei. It seems quite likely that the 
spirem completes its cross-segmentation without further change and gives 
rise to the bivalent chromosomes. This method, while varying somewhat 
from the condition in which regular loops are formed, is seen to admit of the 
same interpretation, namely, that two pieces previously arranged ‘ tandem ’ 
in the spirem become approximated to form a chromosome. 
The chromosomes formed as shown in Figs. 17, 18, 19 now enter upon 
the period of shortening and thickening. Some of the pieces are variously 
oriented toward each other, and it becomes evident that many of the forms 
observed at this time consist of the two sides of a loop which have become 
bent together and applied closely, while others are completely broken apart 
or the line of demarcation is quite distinct. The pieces are often twisted 
over each other at this time, but, as shortening and thickening takes place, 
they seem to straighten somewhat and appear more like straight pieces lying 
alongside each other, or as links or rings. The Y and X forms and the 
figure eight are also to be found. 
There is nothing in the spirem of either Pinus or Thuja to indicate the 
relation which the somatic chromosomes bear to each other. Miss King (’07) 
has shown that in Bufo certain chromosomes have the typical ring-form 
which has been commonly observed in all studies of the heterotypical 
chromosomes, while others are of a dumb-bell shape. In that form the 
spirem from which the chromosomes are formed is never found to be split 
longitudinally, and thus all possibility of their origin as demanded by the 
popular theory of conjugation side by side is precluded. Miss King (’07, 
Figs. 28-31) shows very clearly that the chromosomes are arranged ‘ tandem * 
