PROTOPLASM THROUGH THE WALLS OF VEGETABLE CELLS. 
843 
fore, a subject which must be reserved for another occasion. I would only draw 
attention here, to the striking similarity which the arrangement of the protoplasmic 
threads joining the cells of such endosperms as Strychnos, Tamils, or Dioscorea 
presents to the same arrangement of achromatin fibres which accompanies the 
development of the similar structure in Agrimonia Eupatoria ,* and the close 
resemblance of the barrel form, so beautifully shown by Heterospathe, BentincJcia, 
or Lodoicea, to the like form assumed by the fibrilhe between the dividing nuclei in 
such endosperms as Caltha palustris, t or, indeed, in cell division in general. As 
Tangl| remarks, it seems as if the fibrillse persisted during the subsequent cellulose 
formation and deposition. The appearance, perhaps, suggests that such is the case, 
and that the particles of cellulose have been deposited around the threads. It may 
also be noticed that no instance of a reticulate arrangement of the threads has been 
observed. In any case it is apparent, of course, that grave alterations must be occa¬ 
sioned by subsequent growth and increase in size of the cells, but anything certain 
development alone can decide. 
II. Structure of young endosperm cells. —A number of observations were made 
upon the young endosperms of Archontophcenix Cunninghamii , Sctbal umbraeulif era, 
and Rhopalostylis sapida. In all these cases it was found that when the cell was still 
living, as could be seen from the presence of a well-developed nucleus, the connexion 
between the cells was fully maintained, and therefore that communication had existed 
in any case from a very early period (Plate 69, figs. 14 and 15). 
As the cells grow older profound changes take place in the protoplasm, which usually 
result in the death of the cell. In order to ascertain the fate of the nucleus, portions 
of ripening seeds of Archontophcenix elegans and Rhopalostylis sapida were treated for 
twenty-four hours with saturated watery picric, and well washed with alcohol, until 
the yellow colour of the acid had quite disappeared. Sections were cut, which were 
stained with ha3matoxylin, and mounted in dilute glycerine. It was then apparent 
that well stained nuclei were present in the cells occupying the central portion of the 
seed (Plate 69, fig. 15), and as one gradually traced the staining effects from within, 
outwards, it was seen that the nearer the periphery, the less conspicuous became the 
cell-nucleus, until in the outermost layers no trace of a nucleus could be detected ; its 
substance staining less and less, and its outline becoming more and more badly 
defined. Thus it apparently suffers a complete disorganisation. 
Along with changes of the nucleus proceed alteration of the protoplasm. In many 
seeds— e.g., Phytelephas —but little protoplasmic substance appears to remain in the 
cell. Oil very frequently occurs as a cell content, and sometimes is present in large 
quantities, especially in the Cocoinem— e.g., Cocos, Badris, &c. Small crystals may 
also occur, and in such examples as Diplothemium, Syagrus, and Coryphci aleurone 
grains are met with. In the cases which I examined with special reference to the 
* Stkasburger, ‘ Zellbildung und Zellthilung,’ Tafel I., fig. 15. f Loc. cit., Tafel II., fig. 31. 
X Loc. cit. 
