6/6 
Tho day.—On the Histological Relations between 
V. Summary. 
1. The development of the sieve plates and sieve fields in the phloem 
of Salvia and Cnscuta is found to agree in all essentials with Vitis (Hill) 
and Laminaria (Sykes). It is interesting to find the same type of develop¬ 
ment in the comparatively large sieve tubes of the climber Vitis, in the 
minute sieve tubes of a small plant like Salvia , in a non-chlorophyllaceous 
parasite, and in a seaweed. In every case the connecting threads in the 
young transverse wall of the sieve tubes are each bored out to form 
a single slime string enclosed in a tube of callus. 
2. The outlines of the development of the young haustorium in 
Cuscuta are described. The invading haustorium is compared with a brush 
of hyphae ; the central hyphae push into the pith or fuse with the xylem of 
the host; the ones next surrounding these fuse with the sieve tubes; the 
peripheral ones remain in the cortex. The main mass of the originally 
separate hyphae form in the mature haustorium, by lateral fusion with one 
another, a compact tissue. 
3. Each hypha composing the tissue becomes subdivided into a 
number of cells forming a strand. When the tip of the hypha is connected 
with a sieve tube of the host, this strand of cells is developed into a strand 
of short sieve tubes. Connecting threads and sieves are numerous in the 
subdivision walls, but are never formed in the fusion walls between the 
originally separated hyphae—that is, they are only formed in those cell- 
walls which are genetically connected. 
4. The wall at the tip of the invading hyphae becomes more and more 
mucilaginous as it approaches the functional phloem. In the inner layers 
of the pericycle it is sometimes sufficiently hydrolysed for some of its inner 
layers to stain with callus stains ; this hydrolysis spreads till it affects the 
entire thickness of the wall at the tip. 
5. Preparatory to the formation of a junction with a host sieve tube, 
an invading hypha lays itself more or less alongside the sieve-tube wall, so 
that the two come into lateral contact. The mucilaginous wall of the 
parasite where it touches a sieve area is then absorbed, and the naked 
protoplasm of the hypha applies itself to the sieve area of the host. It is 
nearly always with a sieve field, occasionally with a lateral plate, that the 
junction is made ; the junction sieve plates and sieve fields in all cases 
exactly resemble those of the host under normal conditions. 
6. The protoplasm of the parasite is easily shrunk away from the 
sieve area, and is probably never closely fused with the host protoplasm. 
The translocation of food substances from host to parasite would appear to 
be of the nature of passive filtration, the contents of the sieve tubes, forced 
by internal pressure, escaping through the lateral sieve fields into the 
parasite. This arrangement probably disturbs as little as possible the 
