( 416 ) 
longer the case by any means. On the contrary, the remains of the 
connecting threads gradually disappear as if they were dissolved in 
the protoplasm and this process has long been completed when the 
cell-walls successively appear between these nuclei also. 
Also in the divisions of the endosperm-cells ef /riti/laria and in 
the root-tip of Vieta Kaba mainly the same occurrences take place 
although there are some points of difference to which I shall refer 
presently, and although the formation of the cell-wall follows sooner here. 
How this wall-formation takes place has for the present not been 
investigated by Mr. SrPKeNs, but that it stands in no relation to the 
nuclear spindle or to a cell-plate formed by it, is pretty clear from 
what precedes. A cell-plate in the sense of botanical authors does not 
even occur. 
Although the opinion, so generally spread in botany, that in many 
cases the formation of cell-walls is dependent on nuclear spindles, 
may have a certain probability when we only think of the eross- 
divisions of the cells of growing-points and suchlike, it lacks, 
generally speaking, every foundation. For any one knows that the 
formation of cell-walls can in many cases have nothing to do with 
a nuclear spindle. Not to mention all possible cases of thickening of 
the cell-wall whieh do not correspond to the formation of a primary 
membrane, [ will only mention zoospores which, after having come to 
rest, form a wall; plasmolysed protoplasts of Sperogyra and other 
Algae which cover themselves with a new cell-wall; Caulerpa and 
other Coeloblasts, the protoplasm of which after a lesion produces a 
new wall-piece. 
But also in other cases, which resemble more the cell-divisions in 
erowing-points, it is often easy to show how newly-formed cell-walls 
cannot possibly have been formed in the nuclear spindle. I mention 
the antipodal cells, which so frequently are formed projecting inwardly 
in the embryo-sac connected only for a small part of their surface 
with the cell-wall of the embryo-sac; in any case no more than a 
small part of the free wall-surface can have been formed here in a 
nuclear spindle. A corresponding case is that of the U-shaped walls 
in the epidermal cells of the leaves of ferns, by which the mother- 
cells of stomata are formed. More clearly still one sees the same 
thing in the formation of the stomata of Anemia fravinifolia: the 
stomata lie in the middle of an epidermal cell of the leaf and the 
nucleus of this cell is still pressed against the stoma. A nuclear divi- 
sion has taken place here before the stoma-mother-cell was formed 
in the epidermal cell, and between the two cells so formed there 
certainly was a spindle at first. But in the subsequent cell-division a 
