844 
MR. W. GARDINER ON THE CONTINUITY OF THE 
question the cells appeared to be quite dead, and as such they are simply preyed upon 
by the growing embryo. Thus, all the changes which result in their subsequent 
breaking down proceed from the embryo itself. 
Special experiments were made with sulphuric acid, in order to observe its action in 
cases where a continuity was knowm to exist. In fresh living cells treatment with 
sulphuric acid, and staining with methyl violet and glycerine or Hofmann’s blue, 
showed that, although the protoplasm had contracted, those portions projecting into 
the pits still adhered to the pit membrane, and that the threads of protoplasm running 
through the pit membrane were continuous on either side witn the above-mentioned 
symmetrically opposite processes (figs. 14 and 15). The processes, in fact, appear to 
be held to the pit membrane by the threads in question in all cases where the con¬ 
tinuity is pronounced. Under a low power the individual threads could not be 
distinguished, and the appearance then presented was that of two darkly-stained 
threads united by a lighter-stained area running between them—in fact, the very 
appearance presented by Mimosa and Robinia* (cf. figs. 5, 7, 14, and 15). In the 
case of ripening seeds, the protoplasm may be made to contract slightly from the 
membrane, and then a similar phenomenon is induced to that which occurs in Amicia 
and the bast cells of Mimosa (cf figs. 1 1 and 13), although in them it is not occa¬ 
sioned by loss of vitality, but rather from the fact that the threads are probably 
extremely fine and the continuity not so pronounced as it is in the case of the 
parenchyma cells. In fully-ripe seeds where the cells are dead, the protoplasm always 
contracts away from the cell-wall, and a similar state of things usually occurs when 
the cell has been killed by the action of reagents (see figs. 16 and 22). Thus, both my 
method and my results have received very satisfactory confirmation and elucidation. 
III. General results with ripe endosperms.— As a rule, most of the seeds I examined 
were either one or, at most, two years old. I also made use of some museum spe¬ 
cimens, but decided to reject the results I obtained with them, as I had reason to 
believe that in many cases those results were abnormal. As regards their favourable 
or unfavourable character as material for showing the perforation of the cell-wall by 
protoplasmic threads, seeds greatly differ one from another. In the first place, it 
may be stated, as a general rule, that the thicker the pit membrane the easier can 
the threads be distinguished. In very thin pit membranes the observation of such 
threads as may cross it requires great precaution and care; there is nothing, so to 
speak, for the eye to catch upon, and one has to detect a line within a line. It is 
this very fact that causes endosperm tissue to be so favourable for such an investi¬ 
gation as the present one ; for here, not only, as a rule, are both the cells and the 
pits unusually large, but, what is much more important, the pit membranes are 
thick. In many cases, however, this is not the case, and an examination of such 
examples as Manicaria, Mauritia, or Caryota is quite sufficient to prove that the 
* The results with Bomarea also confirm this. 
