59 
in. broad,’ which is larger than any specimen shows or than his 
figure depicts. I think that the measurements are, at any rate, 
extreme. In any case, I feel sure, for my own part, that Rox- 
burgh’s J/. edule is the common species of the North Coromandel 
coast and adjacent hills, which undoubtedly has shining leaves 
and blue-black pulpy fruit. J cannot identity with it any other 
of the species described by other authors, unless perhaps it is the 
M. ovatum, Sm. in Rees Cyclop. in part, as recognised by Triana. 
Most of the speciméns in the Kew Herbarium written up by U. B. 
Clarke as J. edule var. ovata seem to me to belong rightly to M. 
grande, Retz, as was recognised by Trimen. Of the Madras plants, 
therefore, placed by Clarke under M. edule, I propose to accept 
three species, M. edule, Roxb., M. umbellatum, Burm. f., and 
M. grande, Retz (M. edule var. ovata, Clarke) as well as M. moles- 
tum, a little known plant of the Anamalai hills with quadrangular 
branchlets. Clarke’s var. Rottleriana, distinguished by larger 
flowers and longer peduncles, does not seem capable of separation 
as a species, or even as a variety, from M. umbellatum. It was 
collected by Heyne and seems to be the same as Wall. Cat. 4107 B. 
The genus Jemecylon is one which presents considerable diffi- 
culty, especially when one is dealing with dried specimens, many 
of those available being rather imperfect and *‘ scrappy.”’ It is 
a pity that I had not the Madras Herbarium specimens available, 
but when travelling about the Madras Presidency on Forest De- 
partment work in the years 1882-1890 the genus interested me and 
I collected many specimens myself. It may be not out of place 
here to put on record my idea of the geographical distribution of 
the chief species, because they seem to me to occupy fairly well 
defined regions. In the “dry evergreen’’ forests of the East 
Coast from Orissa down to the latitude of Madras the common and 
almost only species is M@. edule, Roxb. In the more southern 
forests, overlapping MV. edule, M. umbellatum is found from near 
the Kistna river southwards to Cape Comorin and westwards to 
the slopes of the Ghats. It passes over into Ceylon which M. 
edule does not, and in Ceylon is found, apparently quite endemic, 
the pretty and well-marked M/. capifeilatum, so well figured by 
Burmann and in Trimen’s Flora. On the West Coast the chief 
species is M. grande, not unlike, but distinct from M. edule, as 
well as other but less abundant species like M. depressum, M. 
Talbotianum, M. deccanense, M. terminale, M. gracile, and n the 
extreme south M. angustifolium. We now come to the hill species : 
the first to appear, in the Deccan hills of Cuddapah, extending to 
Coimbatore and the Nilgiris, is the slender Jf. Lushingtonit. In 
the Shola forests of the Nilgiris and Pulneys the most common 
species is M. malabaricum, very noticeable about Coonoor and other 
places for its bright flowers. We also get the yellow-leaved MZ. 
flavescens on the east and M. molestum on the west, while towards 
Sispara and Naduvatam the handsome M. sisparense is conspicu- 
ous. Away to the north, on the Bengal-Central Provinces border, 
a little known species occurs at high levels which I have called 
M. madgolense. It is to be hoped that Madras botanists and 
forest officers will continue to pay attention to the genus am 
collect specimens both in flower and fruit, and obtain further 
