674 Benson.—Root Parasitism in Exocarpus . 
where the bands are fine the pellicles are often numerous but minute, and in 
a neighbouring cell where the bands are broad the pellicles may be uniseriate 
and relatively large. This phenomenon rather points against a dependence 
of their formation on plastids (Text-fig. 3 a). 
Connexion of the Haustoria with the Mother Root. 
In Exocarpus and Thesium , one of the most striking features of many 
of these haustoria is the somewhat abrupt change in character which is 
noticeable when one passes from the phloeotracheides to the tracheides of 
the neck. The neck tracheides contain no matrix and no granules. They 
are often scattered, so that in certain sections of a series there may be 
no continuity between them and the phloeotracheides. 
It may be useful to note that neither in Exocarpus nor in Thesium 
have I observed any disintegration of the conducting cells of the haustoria 
such as Dr. Barber describes 1 as occasionally occurring in Santalum. 
Continuity is established, as Solms-Laubach 2 described for Thesium, by 
reticulately thickened tracheides—‘ einzelne vielfach zickzackformige Gefass- 
reihen die Verbindung der Gefassbogen mit dem Biindel der Mutterwurzel 
bewerkstelligen.’ But in Exocarpiis the phloeotracheides are not limited to 
two bands ( £ Gefassbogen ’) as in Thesium. It would be along the tracheides 
of the neck that the first water-supply must come to the developing 
haustorium, and it is with the xylem that they are continuous. 
Comparison of the Haustoria of Exocarpus with those of 
Thesium. 
The Thesium haustorium, like those of all other members of Santalaceae 
hitherto described, shows the phloeotracheides in two bands instead of being 
distributed all over the surface of the nucleus as in Exocarpus. Also there 
is no pad of phloeotracheides at the proximal end of the haustorium. The 
granules are present in the phloeotracheides (Fig. 3, PI. LV), but are far 
smaller and less regular than in Exocarpus. In both genera the phloeo¬ 
tracheides may or may not be regularly formed from a meristem. 
Thesium as a Host. 
In three series which had ostensibly been taken through haustoria 
of Thesium it was found that Thesium was the host. In one case the 
attack had been made by another Thesium root, and in the other two cases 
by haustoria of another type (Fig. 1, PI. LV). I have not been able to 
identify the form, but it is obviously built on the plan of those of the 
1 Barber, 1 . c., Santalum album , Part II, p. 20 (and also Plates III and IX, 12). Proc. of 
Camb. Phil. Soc., vol. xiv, Part III, p. 252. 
Solms-Laubach, 1. c., p. 544. 
