Cuscuta and its Host . 665 
secreted by the protoplasm, it is easy to understand that the ferment 
may change other parts of the wall on its way outwards. 
Just behind the fusing tips, cell-divisions are now taking place in the 
hyphae. Fig. 49 represents a transverse section of a hypha which has been 
divided into three cells; the newly formed cell-walls are packed with con¬ 
necting threads, but there are no connecting threads in the outer fusion 
walls. No connecting threads have ever been seen in the fusion zvalls 
separating adjacent hyphae or adjacent strands of cells formed from these 
hyphae. In quite mature haustoria it is often very difficult to make out the 
boundaries of the original strands, but whenever the preparation is sufficiently 
good it is clear that, whereas sieve tube areas are common in the walls sub¬ 
dividing the hyphae, they are not formed on their original longitudinal 
walls (see Figs. 62 and 64, PL L; w — original wall). 
These facts are perhaps the most interesting outcome of this investiga¬ 
tion, since it would appear that here is at least one incontrovertible case in 
which the distribution of connecting threads is clearly coincident with the 
distribution of genetically connected cell-walls, and does not extend to 
walls associated with one another by later growth. 
Behaviour of the wall at the tip of an invading hypha. It has already 
been stated that the tip of an invading hypha is at first covered by a thin, 
very soft and plastic wall with which the protoplasm is closely in contact 1 
The plasticity of the wall in this region is illustrated by Figs. 46 A and 46 B, 
in which hyphae are shown on their way through the walls of cortical cells ; 
the invading element makes a clean hole in the cortical wall, and squeezes 
its way through the hole. It is often much contracted in this process, 
expanding again when it reaches the cavity of the cortical cell; later the 
hole is enlarged and the hypha regains its original shape. At the edges of 
the hole, host and parasite are almost from the first so firmly cemented 
together that it is difficult to ascertain at exactly what point one wall begins 
and the other ends. 2 In preparations stained with Delafield’s haematoxylin 
the common portion of the wall takes the stain more deeply, and is thus to 
a certain extent distinguishable. 
As the hyphae approach the central cylinder their end walls become 
conspicuously swollen and mucilaginous, and at the same time still more 
plastic. Fig. 50, PL XLIX (less magnified than Fig. 46 B), represents two 
adjacent hyphae passing through the pericycle ; the wall of the right-hand 
element is slightly swollen at the tip. 
The thick-walled cells of the pericycle in Salvia must offer some con¬ 
siderable resistance to the passage of the hyphae. Cases have been met 
1 Peirce also suggested that the end walls are not quite solid ; ‘ kept in partial solution by the 
enzyme they are secreting,’ 1894, p. 114. 
2 Cf. Peirce, 1894, pp. 113-14, and Figs. 8 and 9, PI. VIII. It is this firm union which even 
at an early stage prevents the haustorium from being pulled out of the host tissues. 
