i 9 i6] NOTHNAGEL— REDUCTION DIVISIONS 467 



by a gradual thickening (figs. 20, 21, 2ix, y, 22) form a bivalent 

 (fig. 22). 



Each bivalent, therefore, is composed of 2 somatic chromosomes, 

 presumably maternal and paternal, that previously had an end 

 to end arrangement in the spirem, and which may be open either 

 at both ends or at one end. In the latter case a later segmentation 

 will separate the two. So far as any result that is to follow is 

 concerned, the author conceives it to be of little importance 

 whether it be the one condition or the other. Some claim that the 

 bivalents are always open at both ends, which would necessarily 

 be the case were they derived from a paired spirem; but definite 

 cases have been found where the bivalents were continuous at one 

 end (figs. 22, 23), and in fig. 22, lying under another bivalent, will 

 be seen one in which the two arms have not twisted about each 

 other as yet, but are lying more or less stretched out in the cavity. 

 Under such circumstances the bivalent could not have been formed as 

 Gregoire (16), Allen (i), and Yamanouchi (32, 33) have claimed. 



To repeat once more, each arm of a bivalent is necessarily of a 

 double nature, owing to the approximation of the two halves of 

 single somatic chromosomes in synapsis, so that in cross-section 

 a bivalent has a tetrad arrangement. 



FORMATION OF SPINDLE AND DAUGHTER NUCLEI 



Nothing has been found in the literature describing the third 

 contraction or its relationship to spindle formation. 



Lawson (18, 20) finds the weft of fibers, these being trans- 

 formed cytoplasm, outside the nuclear membrane as early as the 

 spirem stage, although in Allium tricoccum they are not visible 

 until segmentation. 



After carefully studying the paper entitled "Nuclear osmosis 

 as a factor in mitosis" by Lawson (20), and then comparing the 

 same with results found in Allium tricoccum, many points of dis- 

 agreement were encountered. As has been pointed out by Farmer 

 (8), Lawson has used the term "permeable membrane" in describ- 

 ing the nuclear membrane, after which he continues to speak of 

 osmotic systems and exosmosis. In this discussion, when speaking 

 of the membrane in this connection, the term "semipermeable 



