FLORA MONTANA FORMOSiE 1(53 



Stamina 4 infra faucem tubi affixa, filamentis brevibus, antheris 

 oblongis. Ovarium globosum 2-lobnm stylum suba^quans 2-locu- 

 lare, locnlis oo-ovulatis. Capsula globosa 6 mm. in diametro 

 aequans rubra 4 valvis dehiscens. Semina numerosa ovoidea J mm. 

 longa laevia minute reticulata. 



Hab. Ganzan, in montibus Morrison, ad 9141 ped. alt., leg. S. 

 Nagasawa, Oct. 1905, (No. 639) ; in isdem montibus, leg. G. Naka- 

 HAEA, Oct. 1906 ; in monte Morrison, ad 8000 ped. alt., Oct. 1900, 

 (No. 1803), et in montibus centralibus, ad 10000 ped. alt., Nov. 

 1906, (No. 1847), leg. T. Kawakami et U. Mori. 



It is not without hesitation that I have identified the present 

 plant with Elmer's Nertera dentata, transferring it to the genus 

 Locjania. There is a httle doubt about this being similar to Mr. 

 Elmer's. In his diagnosis, the fruits are described as follows: 

 " berries shining vermihon red, 2 lin. in diameter ; globose, frequently 



with a short persistent style " But, in my specimens, they are 



decidedly capsules, 5 mm. in diameter, opening with two valves, 

 each valve being afterward divided into two. After considerincr 

 his description and my specimens, I have wondered if he had not 

 made an error of regarding the fruits as " berries." Mr. E. D. 

 Merrill affirmed me that my plant is exactly the same as the 

 Philippine plant. In superficial observation, it appears very much 

 to be a rubiaceous plant as Mr. Elmer writes in his paper. Never- 

 theless, the plant has superior ovaries, by which character it should 

 not be regarded as belonging to Rubiacea?, but must be referred 

 to Loganiacege. This species is greatly difTerent from any 

 other species of Logania known to us, and it is perliaps the 

 smallest of all the plants belonging to this genus. The plant 

 agrees with the generic characters of Logania in its floral 

 structure, inflorescence and habit, and it is somewhat like the 



