STAMINAL HAIRS OF TRADESCANTIA. 437 



(4). Five minutes later, the two nuclear halves are a noticeable 

 distance from one another (5). All the daughter segments do not 

 always separate from one another simultaneously ; many remain 

 behind and then hasten after the others. At the same time we 

 see the daughter segments, during this movement, bend at the 

 poles, becoming somewhat shorter and correspondingly thicker 

 (5). Between the nuclear halves remains a substance, transparent 

 as glass, which is quickly increased in quantity by the immigra- 

 tion of the plasmic masses previously collected at the poles (5 

 and 6). In this transparent central mass a finer structure is 

 not noticeable, but we can make out later that it is in fact 

 differentiated into threads. It assumes gradually a barrel shape. 

 From twenty-five to thirty minutes may have elapsed since the 

 commencement of the separation, and we see appear in the 

 equatorial plane of the central mass dark dots arranged in rows. 

 In the next moment these dots fuse together, and we notice in 

 their place a sharply defined dark line, a new membrane which 

 has therefore proceeded from the small granules. These latter 

 form what we distinguish as the cell-plate, In the median 

 transparent protoplasmic substance, and at like distance from 

 the two nuclear halves, a cell-plate is therefore first produced, 

 and from this proceeds a separating membrane. Within this, 

 however, is immediately afterwards developed a partition wall 

 of modified cellulose. If the central, barrel-shaped plasmic body 

 has been formed so broad that it fills the entire cross section of 

 the cell, we see the partition wall at once joined on all sides to 

 the wall of the mother-cell. If, on the other hand, the plasmic 

 mass does not occupy the entire cross-section, it in all cases 

 impinges on one side of the wall of the mother-cell ; and we see 

 it, after the young partition wall has been formed on this side, 

 move about inside the cell, so as gradually to come into contact 

 in all directions with the wall of the mother-cell, and thus com- 

 plete those parts of the edges of the partition wall which are still 

 wanting. The central body withdraws therefore as far as is 

 necessary from the partition wall which is already present, and 

 completes the parts which are wanting by forming further cell- 

 plate sections (7-9). During these processes we see the daughter 

 segments bend at their equatorial ends towards the middle line 

 of the nucleus (7, 8). The ends of these segments in this way 

 come ultimately into lateral contact and union. The nuclear 

 segments in the daughter nuclei now again begin to become finely 



