May 1, 1903. J THE TROPICAL AGRICULTURIST. 729 
in length. It waa observed, however, that these 
branchlets were uovv unlike the fastigiate acd alternate" 
leaved biHUches in other respects. Instead of beiug 
decidedly thicker at their bases and tanering upwards 
like the hitter, the opposite-leaved horizontal branch- 
lets remained of uuitorm thickness throughout ; they 
were moreover rather markedly swollen at the nodes, 
which is not the case with the alternate-leaved 
branches ; moat marked difference, perhaps, of all, 
the redness of the youngest part of the shoot, 
instead of deepening witn age into a rusty brown, as hap- 
pens wiih the alttrnate-leaved branches, passed in tbete 
opposite-leaved branchlets first into green and ulti- 
mately into greyish white. After an interval of two more 
seasoiiti a sli ikiug occurrence took place. Just above the 
oldest set of horizontal branchlets, now in their fourth 
season uf growth, a new series ot buds appeared which 
developed into a new set of branches, so that now in 
each axil were to be found two superposed axillary 
twigs. But every one of these new ai^d accessory 
br,»nches oitltired from the first from its subjacent 
oppoaile-leaved companion brauchlet in haviug dark- 
red baik. with alternate and red-veined leaves, in 
being thicker at the base than it is above, and in 
growing fastigiateiy instead of spreading horizon- 
tally. I'his second set of branches was in fact a 
repetition of the system of branches that appeared 
originally in the cuttings. 
Since then the same phenomena have repeated them- 
selves, the branches last described have developed 
their own primary series of horizontal branchlets and 
are now beginning to emit their accessory series of 
fastigiate branches. Moreover, the primary horizontal 
branchlets of the second series are now in full 
flower aud are setting fruits. 
While the cycle described is the normal sequence 
of events it must not ba supposed that it exhibits 
absolute regularity. To begin with, every leaf on a 
vertical alternate-leaved branch does not have a 
horizontal, opposite-leaved branchlet in its axil ; most 
of tuem have, but not all. Again, by no means does 
every branchlet with opposite leaves become ac- 
companied by a superposed vertical alternate-leaved 
branch. N -t infrequently these opposite-leaved branch- 
lets die and are sued before the corresponding vertical 
alternate-leaved branch has appeared. The buds of 
these alternate-leaved vertical branches, iu spite of 
their greater vigour of growth when they do ulti- 
mately burst, often never show at all or lie long 
Utent, ana appear low down on the stem, long after 
the leaves and the horizontal branchlets tnat cor- 
respond have disappeared. Bat where a vertical al- 
ternate-leaved branch does this break, it is at once 
evident that just below its position there have 
formerly been a horizontal branchlet and a leaf. 
Again, though it is usual for only one vertical branch 
to accompany a horizontal brauchlet, now and then 
two or even more than two branches arise at the 
same point. Wnat, however, is certain iu the case of 
Qat is that there are two branching moods, and that 
luese moods alternate. Leafy branches of de iques- 
oent type with alternate-leaves Dear first a set of 
axillary branchlets of excurrent type with opposite- 
leaves, 'ihe latter are comparatively sauic lived 
and do not again branch vegetatively but ultimately 
bear in their axils cymes of flowers. Later the leafy 
alternate-leaved erect branches produce, just above 
the branchlets a new series of branches which 
repeat the erect habit and the alternate leaves of 
the branches that bear them and in turn go through 
the same cycle, 
The facts of the case explain all the divergent 
accounts of the authors wno have been cited. In 
Herbaria only flowering or fruting specimens as a 
rule find a place, hence as a rule only opposite- 
leaved twigs occur in collections. Tnis expl.tuio the 
Btalemeiiis of Hochstetier and of the Genera Flan- 
taruni. Tne writers who say '■ leaves opposite or 
altornate" are more accurate no doubt, but tuey leave 
the tcud couilitiou cuiirely uud^^pUiujd. Iu lae aj.j<> 
of Lindley's figure it can only be supposed that hia 
artist had twigs of both kinds before him, and not 
grasping the significance of their differences had, to 
save space, combined their attributes. Loesener is 
much more nearly exact, and when he says '' leaves 
on sterile branches often allernaie," he is no doubt 
accurate, since all the opposite-leaved branchlets do 
not have flowers. He d<.es not, however, bring out 
the fact that the branchlets with opposite leaves 
so far as we know, are the only ones that ever do 
bear flowers and fruits. Forskael's account, in the 
light of actual facts, now becomes clear. His large 
branches are the branches of the alternate leaved 
deliquescent l}pe; his green branchlets are those 
of the opposite-leaved excurrent g type. The expres- 
sion 'articulate' is not strictly accurate; ihese 
branchlets aie swollen at the nodes it is true, but 
they are not jointed ; nor is he accurate when he 
speaks of them «s annual, for though they do appear 
as a new crop tach year, and though tliey are shoit- 
lived as compared with the erec . branches, they 
persist as a rule, for more than one season. Some 
of them in the case of a plant under observation 
at the Royal Botanic Garcieu, have persisted for 
5 or 6 ye^rs. But when a branch of the erect 
alternate-leaved type appears Just above a branchlet 
of the opposite-leaved type it grows so much more 
vigorously and rapidly than its companion that it 
seems to rob the latter of its nourishment, with the 
result that all save the leaves of its newest growth 
drop off and in a year or two ic makes no more 
growth, loses the remaining leaves, dies and sooner 
or later snaps off close to the branch from which it 
originated. 
In what has gone before, the writer has endea- 
voured to bring together all that has been recorded, 
and to state what he has beeu able to observe with 
regard to Qat. There are, however, not a few points 
that are still obscure; these it may be well to briefly 
review. 
It haa been explained that the writer haa 
failed to get cuttings from the opposite-leaved 
branchlets to hold, It would be interesting to try 
in some locality more suitable to the plant than 
the Gangetic delta, whether this experience holds 
good elsewhere, and whether it is only from 
branches of the alternate-leaved type that cuttings 
can hi struck. Again, in cuttings from branches 
with alternate leaves it was found that t'le buds 
which broke were all buds that gave rise to branches 
with erect habit and alternate leaves. They were 
supposed by the writer to be axillary ; he now knows 
that they were really supra-axil.ary and that in 
theory a bud of the excurrent opposite-leaved type 
must have existed below each, but never became 
developed. No seeds have yet been obtained at 
Calcutta; those who have seeds at their command 
might observe what happens witli seedlings, and 
record whether the first series of branches that 
appear on these are of the normal primary ex- 
current type, or whether with seedlings, aa with 
cuttings, the first series ot primary excurrent opposite- 
leaved branchleis is suppressed. 
Then as to the name. Some writers give Qaf oi Kaf 
(Bove), Kduha (Treasuiji of Botanij), or "Oiifter (Uatt;. 
In the Gardeners' C/(iO)u'de for lb 13 Lindley publishes 
the explanation suggested by a scholary friend that 
the varient Kar for Kat is merely a mistransliteration. 
It very probably is, but the point has not yet beeu 
authoritativ ely settled. Nor, to come to the scientifio 
nomenclature, is it certain that Calha edulis is the 
correct appellHtioc ; whatever the name may be 
auLhority universally cited for it is erroneous. 
Then, is the plant wild in Arabia ? Botta, we 
have seen, speaks of the plant as wild in Yemen. 
The matter is uonbtful. The question whether Co^ee 
ia or is not wild iu Arabia is one that every traveller 
in the PeuiusvUa hopes to settle. The two matters 
hang together, and as the two plants are oaltivated 
side by side both problems will doabtlesg receira the 
same attention , 
