Cuscuta and its Host . 671 
ceases when an adequate number of connexions has been established 
between host and parasite. 
The formation of single lateral connexions has probably the additional 
advantage that it disturbs to the smallest possible extent the mechanics of 
the sieve tubes. It appears to be necessary that the connexion should be 
established with a host element whose wall is otherwise uninjured, though 
occasionally another element of the same tube may have been pierced by 
an invading hypha. No connexion has ever been found between a parasite 
hypha and a sieve area in the wall of an element which has been broken 
through by other hyphae on their way further into the host That is to 
say, at least one transverse sieve plate always intervenes between a broken 
place in a sieve wall and a junction sieve area. 
It is generally accepted that the substances translocating in the sieve 
tubes are under considerable pressure ; 1 in consequence of this pressure 
their contents flow out laterally through the sieve fields as well as termi¬ 
nally through the sieve plates. In the special case, when the sieve tube has 
been interrupted by the growing shaft of the haustorium, the translocating 
substances will all the more attempt to find some lateral outlet, and, as in 
the normal case, will flow gently through the lateral sieve fields into the 
trap laid for them by the invading hyphae. The actual passage of food 
material would thus appear to be more a passive than an active process, 
due mainly to filtration from an enclosed space under pressure. Under 
these conditions a very close connexion between the parasite and host 
protoplasm in the pores of the sieve area would not be necessary. I have 
no conclusive observations to demonstrate how far the parasite protoplasm 
penetrates the pores of the sieve plate or establishes connexion with the 
slime strings of the host or with the protoplasm enclosing them. The ease, 
however, with which the protoplasm of the parasite shrinks away from the 
junction plate is very evident, and one may conclude from this fact, with 
some degree of probability, that the two are never very closely connected. 
Evidence that this connexion is, nevertheless, sufficient for all requirements 
has been obtained by Mr. Mangham, 2 who has seen in microchemical 
preparations indications of the actual passage of sugars from host to 
parasite. 
IV. General Observations. 
A. The efficiency of sieve tubes . The contrast between the behaviour 
of the parasite hyphae in the cortex and the phloem affords evidence of the 
efficiency of the sieve tubes and sieve areas for conduction. 
1 As shown by the work of Fischer, Sachs, &c. See Mangham, 1910, p. 265. 
2 I am indebted to Mr. Mangham for his kind permission to mention his observations. The 
methods employed were those indicated in his paper in Science Progress, 1911, and by means of 
which he obtained the results communicated to the Brit. Assoc, at Sheffield, 1910. 
Y y 3 
