Sepi 1, 1898.] THE TROPICAL AGRICULTUPJST. 
155 
all stages, from young leaves showing translucent 
spots, to pieces of leaves showing well-formed circular 
blisters, also the farther ssages of the blisters 
appearing hairv (under the lens), and others turned 
quite black. The history of the disease is somewhat 
strikiog. It invariably appears on tea, that has not 
been pruned in the autumn. About April it extends 
to the pruned tea, which has by then come into leaf. 
At first it looks like a minute pink spot, which, on 
being viewed through the leaf, is seen to be surrounded 
by a pale margin. This widens, and the upper sur- 
face of the leaf at this point becomes depressed into a 
circular pit that appears shining and moist. The un- 
der surface looks like a wart of a white, woolly appear- 
ance. These warts, as they enlarge, unite together 
and invade the shoots until the whole of the affected 
parts shrivel up. The woolly surface of the warts 
thus seems to be covered with white filaments, but 
I could never detect these as bearing spores. Shortly 
after this the leaves and shoots tarn quite black, 
and fall to the grocmd. At this stage the tea planta- 
tion looks as if it had been burned. I have seen 
hundreds of acres completely ruined in this 
manner. But in two months or so, new shoots 
appear, and the blight is no: seen again, 
as' a rule, till next spring, and even then spasmodi- 
cally, and where unpruned tea exists. It was very bad 
in the spring of 1S95, and in 1397 I could not discover 
a bush with this blight in the very gardens where, at 
the time of my first visit, all operations had been 
completely stopped by it." 
The view entertained by Dr. Watt as to the fungus 
nature of the parasite proved to be correct, micros- 
copic examination showing it to be an undescribed- 
species of ''Exobasidium," possessing feattires of interest 
from the mycologicil standpoint, more especially in 
the production of a dense layer of conidia which covers 
the surface subsequently occupied by the hymenium. 
The earliest indication of the disease is the 
appearance of translucent spots in the leaf, due to the 
disappearance of the chlorophyll and starch grains ; 
this is followed by a rapid increase in the 
number of ce'ls constituting the spongy paren- 
chyma of the leaf and situated within the 
area occupied by the mycelium of the fungus. 
The conspicnous blisters present on leaves that have 
been attacked for some time, are canned by the secon- 
dary increase in the number of leaf-cells over a limited 
area of the surface being resisted by the healthy un- 
yielding tissnes of the leaf ; hence the abnormal 
growth, stimulated by the action of the parasite, as- 
sumes the form of a blister, being concave on the 
upper, and convex on the under sarfaoe of the leaf. 
AVhen the points of infection are numerous on a leaf, 
the originally distinct blisters grow into each other 
during their development. The mycelium is very 
slender, not exceeding 2^ in thickness, sparingly trans- 
versely septate, and tinged Jiith yellow when 
seen in the mass. It runs between the cells, 
which finally become much distorted and separated 
from each other. After bec:m;ng concentrated in 
clusters between the epidermal cells of the convex 
surface of the blister, on the under surface of the 
leaf, the mycelium ruptures the cuticle and appears 
on the surface of the blister under the form of minute, 
densely crowded clusters of hyphae. When the growth 
of the parasite is very vigorous the hymenium is not 
infrequently formed on both surfaces of the 
blister. Some of these hyphae run out into long, 
sterile' filaments, giving a minutely downv or velvety 
appearance to the blister, when seen under a lens ; 
the greater majority of the hyphae, however, remain 
jbort, and produce a single conidium at the apex. The 
conidia are hyaline, or with a tingj of yellow when 
seen in th-? mass, elliptic with somewhat pointed en^s 
l-septatc, slightly cons tricte J at the septum, straight, 
or sometimes very slightly curved, measuring 14 — 
16x5 — 6u It is not unusual to find conidia germi- 
nating ill sitii,ea.ch cell of the conidium nrodncing 
one slender germ tnbe. Mixed with the conidioohores 
are nume oas ^asidia, but these are not sufficiently 
crowded and compact to form a typical hymenium, 
the surface of the tuft constantly remaining loose in 
texttire, resembling the face of a brush rather than a, 
waxy, compact surface. The basldia are subcjlindric, 
and so far as observed, constantly produce two slender 
spine like sterigmata, although ihe presence of four 
daughter nuclei in some preparations ci basidia 
stained with iodine green, would seem to suggest the 
probability of fotir sterigmata being found in some 
Instances. The spores are hyaline, continuous, glabrous 
ovate-oblong,of ten slightly inaeqtiilateral, -5 x 3^. When 
old, the tufts of hyphae appear to contract a little, 
thus becoming more isolated and distant from each 
other, and giving the hymenium a cracked ap» 
pearance. 
The branches do not appear to be disfigured to the 
sa'.ne extent as the leaves by the parasite. 
PEEYEXTITE ilEASURES.— Eemembering tha 
statement by Dr. AVatt'that the disease •■ invariably 
appears on tea that has not been pruned in the 
autumn,'' it seems almost superfluous to suggest that 
autumn pruning shotild be carried out, unless there is 
some very s:rong reason for not doing so. The removal 
of diseased portions before the spores are mature 
would go far towards preventing a recurrence of the 
disease. Such infected parts should be burned, and 
not allowed to remain on the ground. Spraying 
would not, in all probability, be permissible, other- 
wise a solution of potassium sulnhide (one ounce 
to three gallons of water) would prevent to a great 
extent the spread of the disease, if appaed ac the 
time when it first appears. All known species of 
'Exobasidium are parasites, formisg blisters or galls 
on the leaves and branches of the host plant : the 
flowers and fruits are sometimes also attacked. 
Species belonging to the following genera are known 
to serve as host plants : — " Rhododendron , Yacci- 
nium, Andromeda, Cassandra, Arctostaphylos, Eeaum, " 
Saxifraga, Laurus, Symplocos, Arrhenatherum, and 
liromus." The above list of genera belonging to 
widely separated orders of plants, suggests the possi- 
bility of some ally of the tea plant also serving as a 
host for the " blister blight," and if such exist in tha 
forest adjoining the tea plantations, there is but little 
hope of eradicating the dis>-ase until such nurse-plant3 
are removed from the vicinity. 
The following is a diagnosis of the species 
Exobaaidium vexans, '• Massee." '• Hymenophora " 
innat-a, eS'usa, forma varia, vn'go orbicularia vel 
oblonga, in foliis infestatis builas supra concava* 
infra convexas i l2 mm. dinm. interdnm confluentea 
formantia. " Hymenium ''" le^e, siccitate rimosum, 
initio pallidum, dein state albo-pruiuosnm. " Basi- 
dia " cylindracea 30-35 x 5 C d, 2-sterigmatifera ; ste- 
rigmata aculeata, 3 x 0"5 ^. Sporse " ovato-oblongse, 
contin'^s, hyalinae, glabiffi, tsepe inseqtiilaterales, 
5x3^. Conidia fusiformia, hyalina, 1-septata, ad 
septum leviter constricta, 14-36 x .5-6 /j. 
On living leaves and branches of " Camellia Thea.'j 
Assam. 
Fig. 6, Leaf of tea p'ant, with blisters fTmed by 
blister blight ("Exobasidium vexans"'); nat- size. 
Fig. 7, Section through a blister : x 100. Fig.S. Por- 
tion of hymenium of same, showing numerous conidia, 
a. and basidia, 6, bearing two spores each: x 400. Fig. 9 
Free spores ; If^ 400. Fig. 10, Conidia ; x 400. 
THREAD BLIGHT. 
C' Stilbum uantmi," Massee). 
This very remarkable fungus, while agreeing tech- 
nically with the genus " Stilbum " in the structure 
of the couidiophore, differs very materially in the 
presence of an elaborately branched, sterile stroma 
composed of densely interwoven, slender, sparsely 
septate, vaguely branched hyphae, 2-3^ in diameter^ 
combined to form a thii;. white membrane, which 
is inseparable from the bark cr leaf on which it 
grows. This sterile felted mycelium often forms 
white patches several inches in length on tha bark 
of living branches, and then breaks up into irre- 
gularly-branched slender strands, many of which are 
