670 Tho day,—On the Histological Relations behveen 
It is not, of course, possible to be quite clear as to whether the lateral 
twisting of the end of the hypha always takes place before contact with 
a sieve area is actually made (as appears to be the case in Fig. 68, PI. L), 
or whether in some cases it is after the formation of a junction has begun 
that the hypha continues its growth in a lateral direction in order to avoid 
terminal pressure and not to injure the junction plate. 
However it is brought about, lateral contact is so common that it 
would seem probable that even the few cases in which the extreme tip 
appeared to be utilized, e. g. Fig. 76, PI. LI, are probably to be accounted 
for by the plane of the section passing through the heel of the slipper¬ 
shaped end. Such an arrangement as is seen in Fig. 7 6 would be obtained 
by taking a section at right angles to that of the adjoining element in 
Fig. 77, passing through the left-hand end of the junction sieve area. 
Even in the element drawn in Fig. 77 the extreme tip really extended 
further at another focus for some distance beyond the point shown in the 
diagram. 
These facts naturally raise the question whether, after a junction has 
been formed with a sieve area in the wall of one sieve tube, the hypha ever 
goes on to form a junction with other sieve tubes. As my methods do not 
admit of cutting sections in series, it is not possible to deny absolutely that 
this ever happens. If it does occur, it must be only on the extreme peri¬ 
phery of the haustorium, where the invading elements are less closely packed 
together. There could, however, be no advantage to the parasite in forming 
junctions with a second series of sieve tubes if, in order to do so, it would be 
necessary to make holes in the sieve tubes already attacked, and thus 
probably to destroy their efficiency. It may well be that the further growth 
of the hyphae depends on the composition of the phloem in the host. In 
Salvia the phloem consists largely of sieve tubes, and it would thus be 
almost necessary after the formation of a junction to bore through a sieve 
tube in order to reach others. In other hosts, such as Vitis, where the 
phloem contains much parenchyma, the hypha might conceivably make its 
way from one sieve tube to another by way of intervening parenchymatous 
cells, thus to a less extent disturbing the sieve tubes. No sign of any such 
second junction plate has, however, been found even in Vitis, nor any sign 
of a re-formation of the dissolved portion of the wall of the hypha, nor any 
other adaptation to protect the naked area of parasite protoplasm, such as 
would probably be necessary in the neighbourhood of the disused sieve 
plate. Several hyphae may apply themselves laterally to the same sieve 
tube, but apparently each hypha only forms one such lateral connexion 
during its course. The arrangement would thus be similar to that prevalent 
among the hyphae which form junctions with the xylem. These, having 
made one junction, become lignified, and of course lose the power of further 
growth. We may conclude that the development of the invading hyphae 
