712 Beer. — Studies in Spore Development. II. 
Lundegardh ( 15 ), place no importance upon this contraction and consider 
it to be merely a chance arrangement of the nuclear thread. 
Lundegardh ( 15 ), writing of Trollins enropaeus , says, ‘ Ich will unten 
zu zeigen versuchen, dass die “second contraction” eine ganz sekundare 
Erscheinung ist, die sich in Kernen kurz nach der Spaltung des Kernfadens 
zeigt, und dass ich ihr nicht wie Mottier u. a. eine tiefere Bedeutung fur die 
Entwicklungsgeschichte der heterotypischen Chromosomen zuerteilen kann.’ 
Again, in dealing with the Compositae he further writes of the ‘ second 
contraction ’, * Ebensowenig wie bei Trollius habe ich bei den Compositen 
gefunden, dass dieses Stadium etwas spezifisch Eigentumliches mit sich 
bringe.’ 
Rosenberg, although he has not infrequently noted a second contraction 
in the Compositae, is equally sceptical of the importance of this phenomenon 
in the development of the heterotype chromosomes of these plants. 
In my preparations a * second contraction ’ has made such a regular and 
constant appearance at one particular stage in the history of the heterotype 
division that it seems impossible to regard it as a chance occurrence which 
has no importance. Moreover, it shows such a striking relation to the 
development of the bivalent chromosomes that prejudice alone can induce 
one to deny its significance in this process. I have confined my study of 
this stage to the three species Crepis taraxacifolia (Figs. 12-14), Tragopogon 
prate nsis (Figs. 16, a and b, and 17), and Matricaria Chamomilla (Fig. 18), 
in all of which the contraction is very clearly and regularly shown. 
During this stage the coils of the spireme have become arranged in 
a number of loops, the ends of which are drawn together at one point 
at which the nucleolus is very frequently situated. In some cases these 
loops are very long and bent round to accommodate themselves to the 
nuclear space. In these cases the ends of the loops can often only be 
seen by altering the focus, or, if the sections are very thin, they must be 
sought for in the succeeding section of the microtome series (Figs. 13 
and 17). In other cases the loops are shorter and their entire extent falls 
within the field of vision (Figs. 12 and 18). A certain amount of twist- 
ing of the arms of the loops about one another can be seen, but this 
becomes more obvious at a later stage. One or occasionally two nucleoli 
occur in these nuclei. 
It frequently happens, especially where the loops are long, that they 
tend to close together, and their sides become closely approximated. In 
such cases the long loops bending through the narrow limits of the nuclear 
cavity bear a striking resemblance to a hollow spireme consisting of two 
parallel parts which here and there separate from one another, whilst at 
other places they merge together. Where the sections are thin and the 
closed ends of the loops have been severed by the microtome knife the 
resemblance between the two stages is still further heightened. As was 
