310 DR. D. D. CUNNINGHAM ON MYCOIDEA PARASITICA, 
tents melt into a homogeneous mass of an orange or sometimes of a green colour, and a 
certain amount of thickening of the cell-wall occurs. The cell-wall and coloured con- 
tents now become separated by a distinct interval. The coloured mass next begins to 
assume a lobed or more or less distinctly cruciate outline, due to the development of 
marginal projections; and these continuing to grow, and becoming divided. dichoto- 
mously, a flattened disk or rosette is gradually formed (Pl. XLII. fig. 17). Whilst this 
division of the coloured contents is advancing, processes are developed from the cell- 
wall, which, shooting inwards, and giving off branches as the division of the contents 
progresses, ultimately coalesce with one another, and divide the disk into a series of cells 
arranged in a radiant fashion. The formation of new cells now continues by the repe- 
tition of the process described above; and as the original cell-wall of the spore does not 
rupture, for some time at all events, but appears to stretch with the development of its 
_ contents, a series of coherent, flattened, cellular disks are gradually formed. Although 
the septa marking out the constituent cells are at first common to their adjacent cavities, 
they ultimately divide into two layers, so as to transform the originally continuous disk 
first into a series of segments, and ultimately more or less completely into radiating 
filaments, which, although remaining closely in contact, may be separated from one 
another, and in some cases even broken up into their constituent cells. These disks 
vary in size from mere points to 0-4 millim. in diameter, or even more in some cases, and 
adhere closely to the surface of the leaves (Pl. XLIII. figs. 3, 6). 
These disks differ from those of the mature plant both in their position, which is 
superficial to, not beneath, the epidermis of the host, and in their much denser structure 
(Pl. XLIII. fig. 4). In order conveniently to distinguish the two sets of disks from one 
another, those formed directly from the germinating zoospores may be termed germinal 
or primary disks. We have next to follow the steps by means of which the primary 
disks come to be replaced by the subepidermal disks of the mature plant. Many of them 
never are so replaced, but, after persisting for some time, dry up and disappear from the 
surface of the leaves, or are utilized by parasitie fungal elements in a fashion which will 
be subsequently described. Were it not for these limits to the spread of the disease, the 
destruction of leaf-tissue would be incalculably greater than it is; for the number of 
primary disks originally formed is often excessive, not unfrequently being generally dif- 
fused over the entire surface of the leaf. It is a matter of some difficulty to follow the 
progress of development in those specimens which pass beyond this epiphytic condition ; 
but a series of horizontal and vertical sections shows the process to be of the following 
nature. Some of the cells of the primary disk, in place of merely growing outwards 
by the formation and separation of peripheral lobes, give origin to buds from their 
under surface, which gradually penetrate the thickened epidermis, and ultimately reach 
the subepidermal space, or, more correctly, the line of separation between the epidermis 
and the subepidermal layer of cells (Pl. XLIII. fig. 7. Having done so, they take on 
an active growth, and, forcing their way along beneath the epidermis, and dividing 
dichotomously as they advance, soon form a mass of radiating filaments. Owing to 
their mode of origin, these filaments never occur in such regular and coherent disks 
as the primary disks; but, as they continue to spread and to become crowded upon one 
