igi6] NOTHNAGEL— REDUCTION DIVISIONS 465 



to be the approximation of whole somatic chromosomes, or pseudo- 

 reduction. 



FORMATION OF BIVALENTS 



With close observation during early spirem, the double nature 

 is still discernible, although it is rare that the two halves separate 

 as found in Allium cernuum (Mottier and Nothnagel 24), and 

 as commonly found in Lilium. The lumpy condition of the thread 

 at this time is due to the larger chromatin collections of the early 

 prophase (figs. 5-9), this appearance being the basis of the state- 

 ments of Farmer and Moore (9), Mottier (22), and others who 

 interpret such as dividing chromomeres, thereby initiating the 

 longitudinal split. This dual nature has been seen in most cases, 

 although it has been attributed to two sources. Those believing 

 in pseudo-reduction in early prophase assert that it is the two 

 spirems that have paired, while Mottier (21, 22, 23), Strasburger 

 (30, 31), Beer (2), Gates (13), Farmer and Moore (9), and 

 Farmer and Shove (10) say that the spirem has split in preparation 

 for the homotypic division. In this latter case, however, the split 

 was not traced to its origin. The two halves soon approximate so 

 closely that the spirem appears as a homogeneous structure (figs. 

 15, 16). Comparatively few ends are seen when sections are cut 

 12 ju thick, and from all of those observed it appeared to be due to 

 cutting. So far as the stages to follow are concerned, it would 

 make little or no difference whether the spirem be continuous or 

 non-continuous. 



At first, as formerly stated, the spirem is irregularly placed 

 within the nuclear cavity (figs. 14, 15), this being followed by a 

 peripheral arrangement, thereby forming the typical hollow spirem 

 as described by Mottier (21, 22), Mottier and Nothnagel (24), 

 and Beer (2) ; at which time the thread may be followed for a 

 considerable distance. As it thickens and shortens, the second 

 contraction period is entered upon, although Overton (26) fails 

 to find such a stage in Thalictrum purpurascens, Calycanlkus 

 floridus, and Richardia africana; also Gregoire (16) fails to 

 observe it at times in Lilium speciosum; both regarding such an 

 act as of little significance in the reduction division. The author 

 believes this step to be of considerable importance. 



