304 Sykes. — Anatomy and Histology of Macrocystis pyr if era and 
In surface view (Fig. 2, PI. XIX) a few threads, not more than six or 
seven, are seen in each cortical pit, and in longitudinal section (Figs. 3 and 
4, PL XIX) these threads may be seen as very short lines which cross the 
pit-closing membrane and unite the protoplasm of adjoining cells. These 
connecting-threads were seen in old plants in the pit-closing membranes 
in the transverse and longitudinal walls of both inner and outer cortical 
cells. 
b. Hyphae. Numerous single fine threads were demonstrated in the 
cross walls of the anastomosing hyphae in the young material (Fig. 24, 
PI. XIX), but in the older material they are not visible, and it seems probable 
that the great swelling of the walls has stretched and broken them. It 
is impossible to be certain which of the cross walls are due to secondary 
fusions among the hyphae, and one cannot, therefore, say whether threads 
are developed at places where secondary fusions have occurred. To clear 
up this point an examination of young plants will be necessary. 
iii. Development of the sieve-plates and obliteration of the sieve-tubes . 
a. Primary pith filaments. Owing to the absence of sufficiently young 
material the development of the cross walls of the primary pith filaments 
could not be followed, since callus was already present in nearly every case 
investigated. In one instance (Fig. 2 < 5 , PI. XIX) knobs were seen on 
either side of the cross wall, but it was impossible to be certain that they 
were connected by threads, as this stage was observed in spirit material. 
Callus is very soon formed, and Fig. 28, PL XIX, represents a very 
commonly occurring early stage, in which callus-rods are found traversing 
the sieve^plates and a callus-pad has also been already laid down on either 
side of the plate. There seems no doubt that the mass of callus shown here 
and in the later figures is laid down by the protoplasm, and is not formed by 
an alteration of the wall already present, for the original wall, undiminished 
in thickness, can still be seen, and it does not seem possible that such a large 
amount of callus could be produced merely by the swelling of an im- 
measurably small outer layer of this wall. 
Fig. 29, PL XIX, represents a case in which the callus is accumulating in 
the centre of the plate, while Figs. 30, 31, PL XIX, are drawn from older stages 
in which a larger amount of callus has been formed. The callus usually 
accumulates much faster on the side farther from the apex. In Fig. 31, 
Pl. XIX, the protoplasm has begun to lay down callus throughout the length 
of the tube. It is formed first as a thin line on the edge of the protoplasm, 
but soon becomes irregular in its distribution, being laid down in small 
wedges or blocks, adjoining especially dense portions of protoplasm. These 
wedges (cf. Fig. 23, Pl. XIX) meet across the tube in places and nearly choke 
up the channel, but for some time a thin layer of protoplasm is left lining 
