1923] SCHNEIDER, NOTES ON HYBRID BERBERIS 197 



ramulorum fertilium lanceolata vel oblanceolata, apice acuta, basim versus 

 in petiolum brevem attenuata, satis argute graciliter serrata dentibus 

 0.5-1.5 mm. longis utrinque 6-10(-12), superne vivide viridia, nitida, subtus 

 vix vel paullo pallidiora, utrinque laxe elevato-reticulata, minimis exceptis 

 1-2 (-2. 3) cm. longa et 3-6 mm. lata; folia ramulos novellos fulcientia 

 surculorumque late lanceolata, grossius serrulata (serraturis ad 3 mm. 

 longis), ad 3.3 cm. longa et 1 cm. (dentibus exceptis) lata, subtus saepe 

 distincte glaucescentia. Inflorescentiae pleraeque dictincte breviter 

 racemosae vel saltern versus apicem ramulorum fasciculato-racemosae, 

 circiter 6-15-florae, ad 3 cm. longae; pedicelli graciles, 5-8 (infimi ad 10) 

 mm. longi, bracteis lanceolatis acuminatis 2-3 mm. longis suffulti; flores 

 aurei, circiter 6 mm. diametientes ; sepala externa (prophylliformia) parva 

 triangularia vel ovata, media obovato-oblonga, illis duplo majora, interna 



longiora, 5 mm. longa; petala 5-6 mm. longa, 

 obovato-oblonga, apice inciso-emarginata, basi leviter attenuata, glandulis 

 2 ellipticis distinctis separatis aurantiacis praedita; stamina petalis % 

 breviora, apice fere truncata; ovarium ellipticum, stigmate lato crasso 

 sessili, ovulis 2-3 subsessilibus. Fructus elliptici, coccinei, circ. 7 : 4 

 mm. magni, estylares; semina 1-2, cinnamomea, levissime punctata, circ- 

 iter 5 mm. longa. 



Koehne in 1893 apparently described his species from plants in the 



Botanical Garden at Berlin called B. microphylla serrata. In the Kew 

 Garden List mentioned above this form is said to have been received 



M 



a 



serratifoli 



Habitat unknown." Poiret, however, never described such a species; 

 in 1808 he only mentions under B. cretica: "Eadem, foliis dentato-ciliatis, 

 ovatis, obtusis." This form De Candolle in 1821 refers to as "B. cre- 

 tica, serratifolia foliis serrato-ciliatis, Poir. 1. c." Not having seen the 

 specimens of Poiret or De Candolle I cannot elucidate their forms. They 

 may belong to B. nitens which apparently is a cross of B. cretica. In 

 B. provincialis and its var. serrata, however, the stomata in the upper 

 surface of the leaves so characteristic of those of B. cretica are entirely 



absent. 



provincialis 



handsome shrubs densely covered with golden flowers in May. Both are 

 well worth wide distribution in gardens. 



Berberis kewensis, n. sp. vel hybr. 



Frutex ramis elongatis habitu probabiliter B. aristatae; ramuli annotini 

 purpureo-brunnescentes, leviter angulato-sulcati, laeves, biennes cineras- 

 centes vel sordide cinereo-brunnei, vetustiores cinerei; internodia 1.5- 

 2.5 (-3) cm. longa; spinae 3-fidae, rarius simplices, flavescentes, divaricatae, 

 mediocres vel satis validae, mediae 0.6-2.2 cm. longae, lateralibus saepis- 

 sime longiores. 



