26 6 Hill . — 77 ^ Histology of the Sieve-Tubes of A ngio sperms. 
between adjoining sieve-tubes. The sieve-plates occur on the horizontal 
or oblique end-walls of the sieve-tubes, and occasionally on the lateral 
walls also, but their slime-strings are readily distinguished from those of 
the sieve-fields owing to their larger size. As the developmental history 
of the fields appears to be simpler than that of the plates, it may help 
towards a better understanding of the latter structures if the details of sieve- 
field development are first considered. 
Great difficulty has been experienced in trying to demonstrate threads 
in tissues arising from the cambium or in any embryonic tissue, and the 
youngest sieve-tube in which threads can be seen in the lateral walls is at 
some distance from the actual cambium, and removed, in the radial direction, 
by not more than two or three cells from the youngest active sieve-tube. 
That threads have not been successfully seen as yet in the youngest cell- 
walls is due probably to the limitations of our methods rather than to their 
absence. The walls of the very young elements cut off from the cambium 
are extremely thin, and of course contain a high percentage of water, so 
that it seems likely that they undergo relatively violent alterations on 
fixation, and that the demonstration of connecting-threads is thereby 
prevented. Similar distortion of cell-wall structure appears to take place 
with the fixation of mature tissues in Alcohol, since if this reagent is used 
even with full-grown walls it is impossible as a rule to see anything of the 
finer histological details. 
It is not until the walls of the young elements show a distinct and 
definite increase in thickness, somewhat like that described for the phloem 
of Pinus} that it is possible to stain the fine protoplasmic threads. In the 
case of Vitis a differentiation of the lateral walls of the sieve-tubes is soon 
noticeable, giving rise to shallow pits separated by thicker portions of the 
cell-wall. In a transverse section these pits are often seen to be considerably 
elongated in the transverse direction, and in a surface view they are more 
or less elliptical or fusiform in outline, and are closely crowded over the 
surface of the wall. 2 The thin pit-closing membranes are provided with 
delicate, faintly-staining threads, which in a surface view are seen to be 
distributed more or less evenly over the surface. The threads appear to 
be continuous across the membrane, and there is no evidence of any break 
or node at the middle lamella (Fig. 37, PI. XVIII). 
In a slightly older sieve-tube there is a noticeable increase in the 
thickness of the wall, leading to a greater distinction of the pitted areas ; 
the threads, which still present an unbroken appearance, are somewhat 
longer and more prominent than before, but show a tendency to become 
arranged in small secondary groups within the limits of the original pit. 
On examining a pit such as that just described, after the action of Water 
1 Hill, Histology of Sieve-Tubes of Pinus. Ann. Bot., vol. xv, p. 587, ‘The boundary layer.’ 
a Cf. Wilhelm, 1 . c., Figs. 29 and 38. 
