658 Tho day.—On the Histological Relations between 
rod, and a median node is often visible at the middle lamella (Fig. 9, 
PI. XLIX). In many cases each depression becomes a compound pit, and 
the group of callus rods is separated into smaller groups, each group in 
a small pit, or even each separate rod in a pit of its own (Fig. 9). Callus is 
deposited later over the heads of the groups (Figs. 10 and 12) and may occur 
in considerable quantity in the old tubes (Fig. 13). 
The sieve fields of Salvia are of especial interest in connexion with the 
current research, since it is found that it is with these sieve fields that the 
parasitic cells form a junction in order to obtain food materials from their 
host. No cases of connexion by means of the terminal, and few by lateral, 
sieve plates have been demonstrated. 
(b) My preparations contained numerous well-stained examples of the 
developing pits in the walls of both the pericyclic fibres and the lignified 
elements of the xylem. Three such pits in a young pitted vessel are shown in 
section in Fig. 14; groups of threads are seen to traverse the pit-closing 
membrane, connecting the pit fillings. In Fig. 15 the pit-closing membranes 
are seen in surface view, each traversed by a small number of threads. 
II. On the Histology of Cuscuta. 
(a) The walls of the cortical cells of all the species of Cuscuta examined 
were traversed by numerous connecting threads, generally arranged in small 
groups, each group being confined to a pit, and containing some 6-20 
threads, or fewer in the case of the transverse walls. In many cases the 
pits were again divided into smaller pits (Figs. 16, 17, 18). 
(b) The sieve tubes of several species of Cuscuta have been to some 
extent described by Mirande, 1 and my investigations have confirmed his 
account of their gross mature structure as he saw it when they were isolated 
by maceration. The sieve tubes of Cuscuta are much larger than those of 
Salvia . 
As in Salvia , however, the development of the sieve plates, in the two 
species which I studied carefully, agrees in all essentials with Hill’s descrip¬ 
tion of the development of the sieve plates in Vitis , 2 and is also closely 
comparable with that of the sieves in the sieve tubes of Laminaria. 3 
Except when otherwise stated the following account, and also the greater 
number of the figures, are obtained from Cuscuta reflexa. 4 
The terminal sieve plates. In the two species examined the terminal 
sieve plates are simple, each pit containing a single thread or callus rod ; 
groups of threads or callus rods in each pit are, however, occasionally 
found in the terminal plates of the sieve tubes in the haustorium of Cuscuta 
europaea on the Vine. 
1 Mirande, 1898. 2 Hill, 1908. 3 Sykes, 1908. 
4 According to Solereder, 1908, pp. 1003-5, there are sieve plates in the longitudinal walls of 
the sieve tubes of the Monostyleae only. But I have found no difference in this respect between 
C. reflexa (Monostyleae) and C. europaea (Distyleae). 
