668 Thoday.—On the Histological Relations between 
compound sieve plates, but are derived from the host wall only, and are 
utilized in their original condition by the parasite for its own convenience. 1 
We are at once confronted with the question as to what is the ultimate 
fate of the portion of the wall of the parasite intervening between host 
sieve area and parasite protoplasm, since it does not form part of a ‘ com¬ 
pound sieve plate There are obviously two possibilities: the food material 
may pass through the soft and plastic terminal wall, or the wall may be 
dissolved and the naked protoplasm apply itself to the host sieve area. 
It has already been stated that after the parasite hyphae have pene¬ 
trated into the functional phloem, thick mucilaginous blue-staining tips are 
no longer to be found. Some change in their nature must therefore have 
taken place. When an invading cell is applied to a young or mature sieve 
plate, in nearly all cases an unstained area is found between the parasite 
protoplasm and the sieve plate. Since it is the essence of the methods 
employed in this investigation that they demonstrate protoplasm, slime 
strings, callus, &c., but leave the cell-wall unstained, it is not easy in such 
cases as Figs. 76 and 77, PI. LI, to decide whether this unstained area (between 
the parasite protoplasm and the sieve area j) be wall or space. It was not 
found practicable to use a wall stain conjointly with safranin and London 
blue, &c., and in preparations in which wall stains were employed by them¬ 
selves it was very difficult to locate the junction sieve areas. In a few cases, 
however, this method led to useful results. Fig. 72, PL L, is an example of 
a junction plate between parasite and host, stained in Delafield’s haemato- 
xylin. Between the parasite protoplasm and the sieve field there is a large 
unstained area (sp) which is here undoubtedly a space, and must have been 
caused by the solution of the parasite wall in this region. x t x mark the 
termination of the unaltered portion of parasite wall. In Figs. 76 and 77, 
PI. LI, also similar knob-like endings are labelled x> x and suggested 
a similar interpretation even when examined without a wall stain. The 
same preparation was later stained in haematoxylin, and is drawn in Fig. 78, 
where the relations of the walls are clearly demonstrated. 2 It was then 
quite clear that the unstained area here also is a space, and that the 
intervening parasite wall must have been absorbed. 
In Fig. 78, at R, R, are shown small remnants of the parasite wall which 
have apparently not been dissolved with the rest. 
The formation of a space on the parasite side of the sieve area is no 
1 It appears to be always with the longitudinal wall between two sieve tubes that junctions are 
effected ; at any rate in Salvia no satisfactory cases were observed in which a junction was made with 
the wall between a sieve tube and a companion cell; in Vitis there was some evidence that such 
junctions may occasionally be developed. 
2 In such preparations, stained with haematoxylin, after having been examined by the usual 
methods for demonstrating the structure of the sieve areas, the walls stain quite sharply, notwith¬ 
standing the processes employed to slightly swell them, and the termination of the parasite wall 
round the hole which is formed over the sieve area is perfectly clear. 
