72 
Walter Gardiker. 
demonstrated, that a continuily of the protoplasra belween adjacent cells is 
established, in pilted cells by means of protoplasmic filaments travcrsing 
tbe pit-elosing-mcmbrane and in unpitted cells bv means of filaments tra¬ 
versing the thickuess ol the walls. In certain instances thc filaments may 
traverse botli the walls, and the pit-closing-membranes. 
These delicate threads or filaments, like the cells themsclves, may 
present all possible modifications, as lo size, as lo destribution and as to 
strueture. Seen with the grealest ease in such cndosperms as Thrinax, 
Bentinckia, Stevensonia, Latania, Ilowea, Ilelerospathe and Lodoicea, they 
arc less clearly demonstrated in Sabal, Manicaria, Mauritiu, Phoenix, Euterpe 
(Areca) and in most of the Cocoineae , while in Bomarea and Iris little more 
than a striation can be made out, and in very many of the thin-walled 
endosperms merely a staining of the pil-membrane. The threads may be 
thick as in Ilelerospathe, Lodoicea and Bentinckia, or thin as in Manicaria 
Dypsis, Oncosperma and Kenha, and similarly they may be many or few in 
number. 
The filaments obviouslv traverse the pit-closing-membrane, or the 
general cell-wall bv means of delicate perforations in these struetures, and 
thus in the ease of pitted cells the closing membranes are exactly comparable 
to a sieve-plate, while in unpitted cells the w hole area of (he cell-wall must 
be regarded as a giganlic sieve-structure. Allhough in active living cells 
these fine channels contain living protoplasm, yet as the cells die the same 
changes take place in the contents of these channels as in thc contents of 
the cells themselves. Thus in some instances, e.g. ripe seeds of Phytelephas, 
on account ol the general scantiness of the protoplasra, the channels are 
almost empty, and contain only a few granules which stain bul slightly with 
Jodine, and dissolve readily in Sulphuric acid. In Ilelerospathe in the same 
way the protoplasm has become so altered that although it will stain quite 
well w ith Jodine, it colours w ith difficulty—if at all w ith Picric IIoffmann’s 
blue, while in other instances where the modification is not so great e.g. 
Bentinckia, Kentia, Lodoicea the same well-defined blue staining of the 
threads occurs, which takes place equallv in all normal protoplasm. 
Having thus dealt in a somevvhat general manner, with some of the 
charaeteristics common lo endosperm cells, and the threads which traverse 
their walls, I w ill proceed to describe a few special examples in order to 
give some idea of the strueture that a treatment with Chlor. Zine. Jod., and 
Jodine or Picric IIoffmasn’s blue, brings into view. Latania Loddigesii may 
be taken as a typieal example of an endosperm in which communication 
between adjacent cells apparently takes place through the pits only. The 
protoplasmic threads which can be demonstrated eilher with Jodine, or 
with Picric Hoffman.n's blue are plainly seen to traverse the pit-closing- 
membrane. 
The w hole figure of the thread-complex as seen in longitudinal section 
