314 Sykes. — Anatomy and Histology of Macrocystis pyrifera and 
square-ended, more or less rudimentary outer sieve-tubes, to the elongated, 
swollen-walled sieve-tubes of the pith. In the younger sieve-plates be- 
longing to the latter, the threads are rather difficult to demonstrate. In 
favourable material they were, however, well-stained, and surface views, 
showing their arrangement, are given in Figs. 28, 29, PI. XX. Sometimes 
the threads are almost evenly distributed, sometimes they form small groups 
of ten or twenty, arranged in various ways, similar to the methods of 
distribution of the threads in the rudimentary sieve-tubes. 
In a very few of the older sieve-plates these groups of threads have 
become converted into sieve-fields. Each group does not, as in Macrocystis, 
give rise to a single slime-string traversing a single hole in the sieve-plate, 
and enclosed in a callus-rod, but each thread of a group remains inde- 
pendent and forms a separate slime-string surrounded by its own callus-rod. 
The number and distribution of the slime-strings in such a stage as Fig. 32 a, 
PL XX, are found to be similar to those of the threads in P"ig. 29, PI., XX. 
The first step towards the formation of the slime-strings is shown by 
the deeper staining power of the threads when treated by the Safranin 
method. At this stage, if the section be placed in London blue, a faint 
blue ring is seen, both in surface views and optical sections, to surround 
some of the threads (Fig. 30, PI. XX). After this, perforation soon follows, 
and Figs. 31, 32, PL XX, represent surface views of sieve-plates riddled with 
holes. Each thread has become converted into a slime-string running 
through a hole in the plate, but in many cases the slime-strings have fallen 
out in the course of preparation of the section. Each hole is surrounded by 
callus which in surface view appears to spread over the sieve-field areas, 
but does not extend over the portions of the plate between the areas ; 
Fig. 32 b , PI. XX, is a small portion of Fig. 32 a more highly magnified, and 
showing the limits of the callus-patches. In optical section each slime- 
string can be seen in the thickness of the plate to be surrounded by a ring 
of callus, which is of course a tube in longitudinal section. The heads of 
the callus-rods soon fuse and form a small aggregation over the ends of 
each group (Figs. 36, 38, PI. XX). Such aggregations are seen in surface 
view in three places in Fig. 31, PI. XX. 
At this stage the sieve-plate is at its fullest development. The sieve- 
tubes nearer the centre of the medulla, some of which represent the primary 
pith filaments, are in various stages of obliteration. Callus soon begins to 
accumulate over the original patches and from there spreads over the whole 
surface of the sieve-plate (Fig. 37, PI. XX). A large mass of callus is found 
in the innermost elements and is generally much more developed on one 
side of the plate, probably, from analogy with Macrocystis 1 , that side which 
is furthest from the apex (Fig. 39, PI. XX). Not more than two sieve-tubes 
in this state are usually present in any radial section. In some of the 
1 p. 302. 
