322 MERRILL. 



2. Quercus jordanae Laguiia Apiintes Sobre Nuev. Itoblc <1(; Filij). (1875) 

 7, cum lamina; F.-Vill. Nov. App. (1883) 208; Nid. \Wv. IM. V:isc. Filip. (188(i) 

 264; Ceron Cat. PL Herb. (1892) 105. 



Q. vidalii F.-Vill. Nov. App. (1883) 2(1!); Nidul Siiiopsis Atlas XLl (1883) 

 /. 92, f. B.; Ceron Cat. PI. Herb. (1892) 10-4. 



Q. caraballoana F.-Vill. Nov. App. (1883) 209; Vidal 1. e. 265; Ceron 1. c. 165. 



Q. havilandii Von Seem, in Perk. Frag. Fl. Pliilip. (1904) 42, non Stapf. 



Q. sundaica Merr. in Philip. Journ. Sci. 1 (1906) Suppl. 41, non lihinie. 



Luzon, District of Lcpanto, Mount Data, Merrill JfoSO, November, 1905 ; Loher 

 .'/SliS: Province of Benguet, Pauai, Bur. Sci. Jf^iOl , JfJ/SO Means, August, 1907; 

 Mount Tonglon (Santo Tomas), Williams 1321, 1365, October, 1904; For. Bur. 

 5009 Curran, August, 1906; Baguio, Lardizabal 7, 1901: Province of Bataan, 

 Mount Mariveles, For. Bur. 1253 Borden, July, 1904; Whitfard 1186, March, 1905: 

 Province of Tayabas, Mount Banajao, For. Bur. 1912 Gurran d- Merritt, November, 

 1907; Elmer 1903, May, 1907. 



Quercus jordanae Laguna, as here interpreted, is a rather variable species, 

 Init after examining the above series of specimens I have concluded that all are 

 referable to one species. Q. jordanae was placed by its author in tlie section 

 Gyclobalanus, but all the specimens cited above are manifestly of the section 

 Pasania. The species as figured by Laguna has relatively broader leaves tlian 

 has Quercus vidalii as figured by Vidal, but the indumentum seems to be nearly 

 the same in both, as well as the shape of the base and apex of the leaves, and the 

 venation. I have seen the type number of Quercus vidalii in Herb. Kew (Vidal 

 611 his), and it is well matched by the specimens from IVIount Mariveles, cited 

 above. His specimen was from the same region as the type of Quercus jordanae, 

 the Caraballo Mountains, in central Luzon. The specimens from Lepanto and 

 Benguet differ from those of Mount Mariveles and Mount Banajao in having 

 somewhat more coriaceous and slightly more pubescent leaves, and rather more 

 pubescent involucres, the scales being also more prominent, but good differential 

 specific characters appear to be lacking. Vidal ISl//, in Herb. Kew, which was 

 mentioned by Stapf in the original description of Quercus havilandii^" as pos- 

 sibly referable to the Bornean species, is almost certainly referable to the species 

 here considered as Q. jordanae, but the specimen is without flowers and fruit, so 

 that its absolute identification will always be more or less doubtful. I do not 

 consider it to be tlie same as Q. havilandii. Quercus caraballoana F.-Vill., to 

 which Vidal refers his No. 618bis, is surely the same as Q. jordanae (Q. vidalii), 

 although the specimen does not appear to be extant, as I could not find it in the 

 Kew lierbarium. F.-Villar's description however applies very closely to the speci- 

 mens above cited, while Vidal " states that it appeared to him to be very close to 

 Laguna's species, giving only some trivial characters by which it could be dis- 

 tinguished. Quercus sundaica Bl., was admitted by me'- on the strength of idon- 

 tilications made l)y O. Von Seemen, but Blume's species is quite ditl'ereut, its leaves 

 liaviiig about 15 pairs of lateral nerves, wliile Q. jordanae lias but 9 or 10 pairs. 

 A siH'cimen in the V. S. National Herbarium, Ijardizuhal 1, was identified at 

 Berlin as Quercus pruinosa Blume, but this is a manifest error, as Q. pruinosa 

 has (juite dilTerent fruits, and differs from Q. jordanae in many other characters. 

 Tlie specimen determined by Von Seemen as Quercus havilandii,^^ Lohcr J/HIS, is 

 not Stapf's species, but is the .same as the otlicr s))t'cinirns fidiii T^rpaTifo and 

 Benguet here referred to Q. jordanae. 



'"Trans. Linn. Soc. Bot. II 4 (1894) :?.?/. pi. IS. f. 1. 

 "Rev. PI. Vase. Filij). (1886) 265. 

 ^' This Journal ^ (1906) >Sf ?</)/)/. 41. 

 "Perk. Frag. Fl. Philip. (1904) 42. 



