Laminaria saccharina . 
3 T 1 
row. In longitudinal section the outermost cells of each radial row are 
seen to be much elongated and are square-ended, but there is a very gradual 
transition between these and the inner elements, which have much swollen 
longitudinal walls, are dilated at the septa, and, in their turn, pass gradually 
into the medulla. The question of the physiological function of the outer 
cells of the primary cortex is a very interesting one, and, as it is impossible 
to separate them from the inner elements, which are certainly sieve-tubes, 
I will say at once that I propose to call them 4 outer secondary sieve-tubes \ 
The use of this term will be seen to be justified by an investigation of their 
histology. 
Though a sharp line cannot be drawn between primary and secondary 
cortex, there is no reason to suppose that elements of secondary origin ever 
become transformed into outer secondary sieve- tubes ; on the other hand 
I believe that these are entirely derived from the primary cortex, and that, 
when that tissue has been used up, no more sieve-tubes are formed \ The 
oldest specimen of Laminaria saccharina which I examined had a stem 
about three quarters of an inch in diameter, and, in a transverse section 
of this stem, there were found radial rows consisting of some ten or twelve 
elements surrounding the medulla. This number agrees with that observed 
in the radial rows of sieve-tubes in the plant figured, which had a diameter 
of less than half an inch. 
The medulla is now a much more elongated oval in shape, and has 
considerably increased in size since the last stage, several of the inner 
secondary sieve-tubes from each radial row having become included in it. 
A few elements, distinguished by their greater diameter and their 
position nearer the centre, probably represent the primary pith filaments. 
While these are always unbranched, the sieve-tubes formed from the primary 
cortex and added to the medulla often have lateral branches which in many 
cases arise at the level of the sieve-plate, recalling those cases in the young 
plant where some of the inner cortical cells gave rise to hyphae opposite 
their transverse septa. The three-armed sieve-plates thus produced are 
very striking (Fig. 36, PI. XX). 
The hyphae in the medulla have formed a much branched system, 
running in all directions in the jelly, and filling up the centre, but they have 
now no free ends. 
Callus is largely developed in the primary pith filaments, and is also 
present in both the outer and inner secondary sieve-tubes, confirming the 
conclusion already arrived at that these elements are to be looked upon as true 
sieve-tubes. The extent of development of the callus was found to differ 
in winter and summer material, but reference to this point will be made 
later. 
1 Cf. Macrocystis , p. 300. 
