ALABASTRA PHILIPPINENSIA, III. 349 
third basaJ pair about 4 mm within the margin, fainter lateral veins 
formed by the second basal vein for nearly the length of the lamina and 
by the basal vein for a shorter distance, cross-veins and finer venation 
also well evident on both surfaces. 
Luzon, Subprovinco of Benguet, Sablan, Bur. Sci. 135U Finix, in flower Nov 
28, 1910. This has its nearest Philippine ally in E, Mtioniana C. B. Rob ' from 
whicii It 18 distinguished by its larger and thinner leaves, of different shape and 
venation. Among outside species, the nearest alliance seems to be with E vale- 
toniana King, which has coriaceous leaves of different color, wiUi more numerous 
nerves. The species is named for the supposed first botanical collector in the 
mountainous parts of nortliern Luzon. 
Eugenia dementis C. B. Rob. in Philip. Journ. Sci. 4 (1909) Bot. 383. 
Comparison of the material upon which this species was based, now sup- 
plemented by For. Bur. 16/(56 Pray, also from Lake Lanao, Mindanao, with 
Eugenia. lanceoJata Lam., as repressented by C. P. 2SGS, from Ceylon, shows 
that tlie two species are very closely allied. The leaves of the type of E. de- 
mentis, Mrs. Clemens 1036, have the veins much more crowded and more evident 
than the Ceylon plant, but on Mrs, Clemens 1113, whicli in spite of other differ- 
ences I am disposed to consider specifically the same, the veins, a_r^ less numerous 
and more widely separated, rather closely approaching those of the hitter. In 
this respect, Fray's collection agrees with Mrs. Glemeyis 1036. Eugenia lanceo- 
late Lam. is considered to be the same as E. wightiana Wight, but the venation 
of the latter, as shown by Wight's plate,"* is much more open than that of 
typical E, dementis. Moreover, it shows no submarginal vein, although this is 
present on C. P. 2863 in practically exact agreement with Mrs. Clemens 1113, 
whereas in Mrs. Clemens 1036 and For. Bur. 16/^56 the vein is almost at the 
extreme margin. Reexamination of the flowers of the type of E. dementis shows 
that the petals are often more than 4, although I have not succeeded in finding 
more than 6. Ordinarily, the outermost is the largest, forming a calyptra, and 
the others are more or less united with one another but separate from the outer, 
the result being that it is rather difficult to determine just how many petals are 
represented. 
Eugenia corymbifera Koord. «t Val. in Meded. Lands Plant. 40 (1900) 99. 
Luzon, Province of Xueva Vizcaya, For. Bur, 10882 Curran: Province of 
Ziimbales, Mount Pinatubo, Loher 60JfO: Province of Bataan, Limay, For. Bur. 
10140 Curran; Duale, For. Bur. 20055 Topado. 
From Koordera and Valeton's description, taken in conjunction with two 
sheets form the Botanical Gardens at Buitenzorg, V, C. 59 and V, B, 25, the above 
identification seems quite positive. The second and third of the above collections 
are between flower and fruit, and have distinctly stouter inflorescence-branches 
than the Javan, but the remaining Bataan plant is in early bud, they are then 
more slender than in the latter. In general, the leaves of the Philippine plants 
are sliglitly different in shape, rather elliptic or elliptic-lanceolate than oblong- 
lanceolate, but so are some on the Javan specimens. The venation is the same, 
but more conspicuous on the Philippine leaves; morever their petioles are 
rather stouter. However, these are quite insufficient characters upon which to 
separate the series of the two regions. Of the six collections above cited, V, B, 25, 
is at least superficially, the most distinct. 
"Ic. Ph: pt. SS9. 
