74 
Walter Gardiner. 
visible. In Auciiba Japonica, and inany other thin walled endosperms there 
ean only be demonstrated a definite staining of the pit-membrane. 
In Strychnos neither Tangl nor myself have been able to delect the 
presence of threads traversing the eell-wall in the layers of cells immediately 
beneath the free surface of the seed, but in Turnus, they ean be seen in all 
the cells equally. Again in Kentia, Howea, and Lodoicea the threads tra- 
versing the general cell-walls could not be made out in the cells situated 
lowards the centre of the endosperip tissue, but only in the more exterior 
layers, while in Bentinckia they ean be demonstrated in a section of auy 
part of the seed. In many instances it is a matter of some difficulty to 
observe that the threads actually cross the middle lamella, but in those cells 
in which that structure is but little developed, such as Heterosputhe, Latania, 
Lodoicea, Bentinckia, Tamus &c the undoubted Perforation of the middle 
lamella is plainly and conclusively evident. In ripe endosperm cells, the 
protoplasm has become so altered, or is so small in quantity, that it fre- 
quently shrinks from the cell-wall and it does not appear as if the threads 
traversing the closing membrane, were really eonlinuous \vilh the proto- 
plasmic processes entering the pits, or in unpitted cells with the general 
protoplasmic body; but if living ceils be taken, and afler having been 
svvollen with Sulphuric acid (which at once kills and fixes the protoplasm) 
are stained with Picric Hoffmann’s blue, it will be apparent that the threads 
do unite with the general cell-protoplasm: that they are eonlinuous with 
the pit processes, and that moreover these processes are iu consequence 
actually held on to the pit membrane for example, in Archontophoenix or 
Bhopalostylis .') 
The threads as demonstrated by Jodine are more distinctly brought 
into view than with Picric IIoffmann’s blue aud they also appear decidedly 
larger in size. The latter phenomena seems to be produced in consequence 
ol the fact that the Chlor. Zinc. Jod. besides marking out and iutensifying the 
staining action of the .Jodine upon the actual threads themsclves, gradualb 
precipitates in virtue of its dein drating properties the Jodine filling the 
capillary tubes. Thus Jodine appears to demonstrate the actual size of the 
channels, in addition to staining the threads which in ripe dead seeds have 
uudergone a definite amount of shrinking, whereas Picric Hoffmann’s blue 
in any case demonstrates the actual size of the threads alone. Moreover it 
appears as if there was even a further aggregation of precipitated Jodine 
around that already known down, for the channels appear abnormally large, 
and consequently the threads abnormally thick and the fact that in sections 
which have been first treated with Chlor. Zinc. Jod., and subsequentlv 
staiuod with Jodine, the threads have appreciably diminished in thickness, 
and appear less strongly defined certainly afiords some evidence that such 
I Tbis completeiy conlirms my results with Mimosa, Kobinia, &c. 
