the Sieve- tubes in Laminarieae. 105 
the callous degeneration, as may be seen from the shading off 
of the red in the figure. 
Regarding the development of the callus in the trump et- 
hyphae ; this begins quite early in their history, and may be 
seen commencing in Fig. 7 in M. zoster aefolia. At this time 
the walls are much swollen up as compared with still younger 
stages. The layer of the wall which first alters, i.e. that 
coloured pink in Fig. 7 , is the outermost one next the middle 
lamella. At first this change occurs only opposite the sieve- 
plate, then later extending a certain distance down the sides 
of the trumpet-hypha. Later still the inner layers become 
similarly altered. In this figure I have not represented the 
perforations as shown in Fig. 8 \_Nereocystis L utheana \ . They 
are very fine, appearing as mere striae. As I have already 
said, the narrow parts of the trumpet-hyphae are sometimes 
involved in the callus-formation. Usually this change passes 
along them centrifugally from the sieve-plate. In one speci- 
men of M. luxurians examined I found that the callus change 
commenced at many points in the narrower parts of the 
hyphae, quite independently — though simultaneously — with 
the change in the dilated portion. This is seen in Fig. 10, and 
is most important when considered in connection with Figs. 7 , 
8 and 9. 
At the regions other than those marked A and A 1 the wall 
has completely broken down into callus ; and by the swelling 
up involved in this change the lumen is obliterated ; 0. /. re- 
presents this lumen dwindled to a mere line. At A callus 
formation has not yet commenced and the hypha is here in its 
normal condition, and a distinct lumen is present. At A 1 
callus-formation has progressed, but the inmost layer is not 
yet entirely converted here. When the whole hypha has 
undergone the alteration it will be no longer constricted, and 
will show no traces of the intermittent manner of the change. 
Leaving now the trumpet-hyphae, I will go on to describe 
the appearances met with in the true sieve-tubes. These tubes 
are fairly wide, running longitudinally in the zone around the 
central hyphal strand. Their average width is about *061 mm. 
I 
