222 BROWN. 
the flower are visible (fig. 8). Owing, however, to the larger 
size of the inner bracts more and more of these become visible as 
growth proceeds (fig. 9). Finally the corolla appears through 
the bracts and the bud looses its rounded shape and becomes tur- 
binate (fig. 10). After this the flower soon opens. The growth 
of the bud is very slow, a bud about the size of one shown in 
figure 9 taking six weeks to develop into a fully expanded flower. 
Peirce ^^ states that both the sieve tubes and xylem of Brug- 
mansia zippelii are connected with the corresponding elements 
of the host. Schaar ^^ describes the xylem of Rafflesia rochus- 
senii as connected with that of the host but was unable to find 
any union between the sieve tubes of the two. The medullary 
rays comprise a large part of the roots of the Cissus on which 
Rafflesia manillana is parasitic. In consequence of this a con- 
siderable proportion of the vascular elements of the parasite 
end in connection with the medullary rays of the host. Both 
the xylem and sieve tubes of Rafflesia are, however, frequently 
in direct union with the corresponding elements of Cissus. Ow- 
ing to the concentric arrangement of the bundles of the parasite, 
this is true of a larger proportion of the xylem than of the sieve 
tubes and in one specimen in which the bundles of the parasite 
appeared to end completely within the cambium of the host, 
there seemed to be little if any chance for a connection between 
the sieve tubes of host and parasite. In the latter case, how- 
ever, the parasite appeared to be as vigorous as usual. An 
examination of figures 3 and 4 shows that a large part of the 
base of the shoot of Rafflesia manillana is surrounded by the 
phloem of the host. This would appear to obviate the necessity 
for a very close connection between the sieve tubes of host and 
parasite, and probably explains the vigorous condition of the 
parasite in the case just mentioned. 
The flowers of Rafflesia manillana, observed, decayed soon 
after opening and gradually disappeared. The part of the bark 
of the host in contact with the shoot also disappeared. The 
vase-shaped mass of xylem around the base of the shoot, how- 
ever, remained as a prominent scar on the root. A recent scar 
is shown in figure 25 and an older one in figure 26. As the 
flower disappears a callous is formed over the scar. 
An examination of figures 27 and 28, representing a cross 
and longitudinal section of a root with a scar, appears to show 
" Peirce, G. J., Ann. Bot 7 (1893) 291-327. 
"Schaar, F., Sitzber. Acad. Wien 107 (1898) Abt. 1 1039-1056. 
