110 ART. I. B. HAYATA : 



cnspidato-acnta vel -obtnsa, ad summum obtusa, basi rotundata. 

 Paniculae terminales 15 cm. longge 6 cm. latse, pyramidales, vel 

 racemose, axillares. Calyces punctati late campanulati obscure 

 5-sulcati, 4 mm. longi, 7 mm. in diametro, 5-lobati, lobis late tri- 

 angularibus 2 mm. longis 3 mm. latis, breve acutis. Petala 5, 

 valvata, angustata 12 mm. longa, 5 mm. lata, apice acuta, margine 

 lamellata.. Stamina 10, 2-seriata, exterioribus longioribus 12J mm. 

 longis, filamentis 11 mm. longis angustatis circ. 1 mm. latis apice 

 2-dentatis, supra dentem abrupte contractis. Antherao orbiculares 

 utrinque retusa3. Styli 5, 11 mm. longi, apice capitellati. 



Hab. Alvö : Tanasliü, leg. Gr. Nakahara, Jan. 1907; Taitö 

 Bokusekikaku, leg. K. Miyake, Dec. 1899 ; Kachinro, Mart. 1898 

 Kötösliö, leg. T. Kawakami et G. Nakahaea, Feb. 1906, (No. 988) 

 Giran : Hachirisha, leg. T. Kawakami et U. Moei, Juni. 1906, (No. 

 1376) ; Taitö : Suibi, leg. T. Kawakami et G. Nakahara, Jan. 1906 

 (No. 670). 



The present plant is very near D. s cabra Thunb. and D. indchra 

 YiDAL ; but it differs from the former in having dentate stamens 

 and in the number of styles which are usually five ; and from the 

 latter, in having quite obtuse leaves. In Vidal's plant, the leaves 

 are usually more acuminate, and very acute at the apex. Deutzia 

 crenata S. et Z. o. Taiwanevsis Maxim, cited above may possibly 

 be identical with this plant, though I have found some different 

 points between the description of Maximowicz and that of the 

 present one ; for the Deutzia credited in Formosa by Maximowicz, 

 so far as is considered from the locality given by him, may not 

 l^e otherwise than this plant as it is the only one to be found in 

 the lowland of Formosa. Maximowicz's plant is not near D. crenata 

 S. et Z., nor is it referable to a variety of that species ; but it must 

 be a quite different one, so far as the diagnosis is concerned. 



