FORMATION OF THE CELL-PLATE AND SEPTUM. 105 



a layer of considerable thickness, is finally consumed by the young daughter-nuclei, 

 which increase in size accordingly. An absorption of the protoplasm of the cell, 

 collected together at the poles of the nucleus, can also be clearly made out in 

 the staminal hairs of Tradescaniia! 



After these changes have occurred in the nucleus and protoplasm, the forma- 

 tion of the new partition walls begins, and I will describe this also according to 

 Strasburger's statements. 'That may be regarded as the commonest form of cell- 

 division in the vegetable kingdom,' he says, 'which is accomplished by means of 

 a partition wall originating in the connecting fibrillae between the nuclei. The 

 few filaments which finally remain behind between the separating halves of the 

 nuclear disc, and which are to be referred to spindle fibrillae, become elongated 

 and increased in number by the deposition of new cell protoplasm, which becomes 

 like them differentiated into filaments. The newly added filaments are not to be 

 distinguished from those originally present : they react like these, and consequently 

 again support the view that the spindle fibrillae are cell protoplasm.' Strasburger 

 understands by the expression ' cell-plate,' a plate-like arrangement of small granules 

 from which the new partition-wall arises. He says it is difficult to decide as to 

 the chemical nature of these granules. 



' So much is certain,' says Strasburger, ' they enter directly into the formation 

 of the cell-wall. There is thus probably not a layer formed of protoplasm, which 

 subsequently becomes split up and secretes cellulose at the separating surfaces ; 

 but the cellulose wall probably originates directly from material conveyed to the 

 spot. I have been able to establish in the living Spirogyra that the granules destined 

 for the formation of the partition-wall wander as such to the place where they 

 are employed. In other (the most numerous) cases, on the contrary, the granules 

 appear to be formed at the spot, there and then. 



' It is especially striking that they are at first small, gradually becoming larger. 

 The connecting fibrillae usually become laterally extended so far, that they spread 

 over the entire transverse section of the cell. Where this has happened, the cell- 

 plate also extends across the whole cell. The cellulose wall is formed simul- 

 taneously from this, and fits close on the wall of the mother-cell at the periphery. 

 Where the complex of connecting fibrillae is not able to extend completely across 

 the cell, it attaches itself first to one side wall of the cell, and, fitting close 

 to this, the formation of the partition-wall out of the cell-plate begins. From the 

 developed wall, however, the complex of fibrillae becomes slowly withdrawn, 

 growing at the same time at its free edges by the continual formation of new 

 connecting fibrillae; and within this the cell-plate becomes completed, until it 

 extends completely across the cell. These differences are conditioned by the 

 size of the lumen of the cell, in relation to the mass of the protoplasm. In the 

 cases where the cell-plates at once traverse the lumen of the cell, or at least but 

 few movements are executed to this end, the cell-nucleus also lies approximately 

 in the middle of the cell : where, however, the cell-plate has to extend across the 

 lumen progressively, the cell-nucleus lies on one side wall of the cell, and becomes 

 divided in this parietal position. Only after the foundation of the cell-wall is a 

 connected layer of protoplasm produced on both sides of it. In this numerous 

 ■granules are often found still embedded, which have not been used up in the 



