1C4 A NEW AND rAlUSITIC
sections of the laminar tissues. In euch aeetione the whole depth of the green
liypodermal tissue is seen to be permeated by - enormous mycelial tubes ( P l a t e V I I I,
Figs. 12—17). Tlie larger of these run moro or less parallel \pith tlîe length of the
axis and are crowded with vacuolate, coarsely granular, colourless or pale ochreoaa
protoplasm exhibiting very active streaming movements. The largest tubes give off
numerous Bomewhat finer branches which are provided with flattened haustorial
jiroceases closoly applied to the outer surfaces of neighbouring cells (Plate VIII,
F i g s . 12b, 14, 16, 17). Here and there branches pass off obliquely towards the
epidermis, and, when tliey reach substomatic spaces, dilate and completely fill these.
One or sometimes two processes are <:iven off from this swelling and force tlieir way
out tliroiigh the stomatic orifice to form secondary dilatations beyond it, and from these
the fertile filaments are in their tura given off (Plate VIII, Figs. 12a, 13). The
filaments become somewhat dilated terminally so as to form a capitulum, and from this a
number of oonidiiferous capitella corresponding to tliose of Choancphora Guminjhamiana are
given off. lû detail, however, the characters of the capitella are very distinct in the two
species. When mature the capitella of the species here dealt with are abruptly truncate
and resemble the conidiiferous cells of Bnlrytk cincrea Pers., rather than the rounded capitella
of Choanephnra Oiinninghamiaiia. Moreover, when the ounidia become detached the
capitella never form the beautiful funnels from which the genus derives its name, which
is no doubt owing to the original difference in their form and to the absence of any
appreciable differences in the thickness of the walls over their lateral aad terminal surfaces.
The number of capitella present in different instances varies considerably, but in
many cases may amount to ten or more. The tnincate extremity becomes covered with
short sterigmata each of which gives orij^àn to a cmulium. The conidia when mature
are very deciduous and fall off much more readily tlian those of Ohomephora Omninghamiana
do, this being partially, uo doubt, owin.ç to the absence of fanuels. In artificial
fulrures in nutritive infusions the conidial fructification comparatively rarely makes its
appearance, and when it does occur is feebly developsd. The cipitula are small and the
capitella few in number, or they may both be completely suppressed and the sterigmata
arise directly from the stem in more or less peronosporoid fashion (Plate I X , Fig. 10).
T i i e conidia when mature have average dimensiens of 0 - 0 1 5 x 0 ' 0 0 8 m.m. They are
fusiform in outline and are provided with a finely longitudinally striate epispore, differing
thus both in form and structure from the conidia of Choanephora Cunniaghamiana
(Plate I X , Figs. 1, 17, 18). This conidial fructification is the only one which 1 have ever
met with on the plant as a parasite on Ipom ea ruhro-uwruka, but probably zygospores
may be produced within the tissues when external conditions cease to favour the production
of conidia. Under favourable conditions, iiowever, during its parasitic existence
the mycelium produces conidial fructification alone and in enormous profusion; that is
to say, it only produces that form of fructification whidi is associated with excessive
nutrition and vegetative growth. Both in regard to tliis species and to Choanephora
Cunninghamiuna the result of artificial cultures appears conclusively to show that nutritional
conditions are the main factors determining the particular form which the fructification
shall assume in individual instancL-s, Where very abundant nutritive supply is
present jjurely vegetative growth occurs; short of this conidial fructification makes its
appearance; with greater diminution of supply the sporangial and zygosporic fructifications
t e n i t j ;ippeav; and, finally, in the presence of minimal nutrition clilamydosporcs aione
a i e p'oduced.
ARIICIT:;» o i ' CITOA.VEPHORA. 165
The species is certainly ca[)ab[e of parasitic exi.stei)ce on other hosts than Ipomcsa
rubro-ccBrulea. The conidia or sporangial .spores when sown on the petals of the common
Zinnia elegana and kept in a moist atmosphere germinate freely, giving origin to germitial
tubes which penetrate the tissues and very rapidly cause their death, producing a crop of
fructificatiim which, although not so vij^orous as that occuriing on Jpomcea ruhro-ccsrulea,
is still of conidial type. In this it differs from Choanephora Cunninghamiana the conidia
or spores of which when sown on Zinnia-flowers fail to givo rise to any blighting of
the latter. As I have more than once observed Zinnias affected by what on casual inspection
I took to be Choanephora Cunninghamiana, there can be little question that these
flowers occasionally play the part of natural hosts to the other species.
The diffusion of the blight in Ipomcca nibro-cceruka .seems to be mainly determined
by direct contact with previously affected tissues, blighted portions of axes serving as
sources of infection of adjoining healthy ones, and the adhesion of blighted and wilted
leaves to healthy ones or to healthy axes being followed by the appearance of the disease
in the latter.
I n artificial cultivations, whether in infusions of Ipomwa ruho ecerttlea or of the
flowers of Hihsciis rosa-sincnsis, the conidial fructification is ordinarily entirely suppressed,
and, when it dees make its appearance in connection with greatly coDContraied
media, it is only feebly developed. Very few lieads of normal type are developed, and
these at utmost bear a very limited number of capitella. In most instances the conidia
arise directly from the primary capitulum. or even tiiis is suppressed, and they arise
simply from the undilated tip of the stem, or at short intervals from one another in its
neighbourhood, so that the fnictifii-ation comes to present an almost peronosporio
character (^Plate J X , Fig. 10). In certain cases only a single terminal conidium may
be present (Plate I X , Fig. 10).
When cultivated in infusions of tiie leaves of Ipomwa rubro-eccrulea or of tha
flowers of Sibiseus rosa-sinensis of moderate strength, or on boiled petals of the latter
planf, sporangic fructification is ileveioped in great abundance, either alone or associated
with varying numbers of zygospores, and it is only in extremely concenlrated infusions,
specially of Hibiscus, that conidial fructification also begins to make its appearance
in association with the sporangic one. The abundance of sporangia produced in cultivations
in media of average quality is escossive. Under normal circumstances they
are produced in far greater numbers and attain a much greater size than those of
Choanephora Cunninghamiana ever are or do (Plate V I I I , Figs. 3, 20; Plate I X , Fig.
13). At an early stage of development they are of an amber colour owing to the tint of
the sporangic membrane, but on ripening they become first brown and uliimately almost
black owing t o the dense accumulation of the brown spores which are formed within
them (Plate V I I I , Figs. 3, 20). The stem is ordinarily abruptly flexed immediately
beneath the origin of the sporangium, so that the latter looks more or less downwards.
On maturing, the sporangia dehisce in valvular fashion allowing the spores to escape,
whilst the two halves of the metnbrane remain attached to the tip of the stem on
either side of the largo cidumella. (Plate V I I I , Fig. 191. The spurangio membrane is •
ordinarily finely tuberculate.
T l i e spores are very peculiar. Like the conidia they are fusiform in outline and
provided with a finely striate brown epispore (Plate IX, Fig. 2). They have average
dimensions of 0-0168X 0 ' 0 0 8 9 ni.m., and are therefore relatively somewhat narrower
than the conidia, and the brown tint of the epispore is as a rule somewhat deeper than
ANN. R o r . BOT. GAED. CALCDITA VOX.. V I .