NOTES AN'D ABSTRACTS. 



529 



Pyrus yunnanensis (Bot. Mag. tab. 8629). — China. Family Rosaceae. Tribe 

 Pomeae. Tree, 20-30 ft. high. Leaves deciduous, ovate, serrate, 2-4 \ in. long. 

 Flowers f in. wide, in racemose corymbs, i \ in. across. Petals pale pink. Fruit 

 globose, i in. diam. ; deep red, specked with white dots. — G. H. 



Rhododendron concinnum {Bot. Mag. tab. 8620). — Western China. Family 

 Ericaceae. Tribe Rhodorcae. Shrub. Leaves 1 J— 3 in. long. Flowers sub-umbel- 

 late, 2 in. across, carmine. — G. H. 



Rhododendron Davidsonianurn {Bot. Mag. tab. 8605). — China. Family Erica- 

 ceae. Tribe Rhodoreae. Shrub, 3^-10 ft. high. Leaves 1-2 £ in. long. Corolla 

 pale rose, i£ in. across ; lobes oblong ; stamens exserted, spreading. — G. H. 



Rhododendron moupinense {Bot. Mag. tab. 8598). — Western China. Family 

 Ericaceae. Tribe Rhodoreae. Shrub, 2^ ft. high. Often epiphytic, wild. Leaves if in. 

 long. Flowers 1-3-nate. Corolla white, 2^ in. across, red-dotted on posterior 

 side. Stamens, exserted with purple anthers. — G. H. 



Rhododendron Souliei {Bot. Mag. tab. 8622). — Western China. Family Erica- 

 ceae. Tribe Rhodoreae. Shrub. Leaves oblong, rounded at apex, 1^-3 in. long, 

 pale green below. Flowers in clusters, about 8-fiowered. Corolla rosy red in 

 bud, white with a rosy flush open, 2^ in. across, 5-6-lobed. — G. H. 



Rhododendron stamineum {Bot. Mag. tab. 8601).— Western China. Family 

 Ericaceae. Tribe Rhodoreae. Shrub. Leaves 2-4 in. long. Inflorescence short 

 10-nate, flowers fragrant. Corolla white, posterior petals yellow at the base, 

 i\ in. across. — G. H. 



Rhododendrons and Lime. By A. Grove {Gard. Chron. Jan. 29, 1916, p. 65). 

 A record of an interesting experiment. A set of seedlings of Asiatic species were 

 planted in August 191 4 in naturally calcareous soil and further supplies of lime 

 added. Some have already died and others are sickly, but many, especially 

 small-leaved species, are still flourishing. Sixteen species are mentioned as doing 

 well.— E. A. B. 



Root-knot Disease. By L. Childs (U.S.A. St. Com. Hort., Cal., Bull. vol. ii. 

 No. 12, December 1913 ; 18 figs.). — Root-knot is usually caused by a minute 

 semi-transparent worm, Heterodera radicicola, which establishes itself in the 

 issue of the root system, producing an abnormal development of the root. At 

 present no entirely satisfactory method of control for orchard or field work has 

 been discovered. In greenhouse and seed-bed steam is most satisfactory. 

 Formaldehyde is a good chemical to apply. Successful treatment has been 

 obtained in using the following formula : 



Commercial (36 to 40 per cent.) formaldehyde . 1 part. 



Water . . . . . . . .100 parts. 



Apphcation per square ya.rd. . . . . . 1 to if gallons. 



V, G. J. 



Rose Aphis, The. By H. M. Russell (U.S.A. Dept. Agr., Bur. Entom., 

 Bull. 90, May 191 4 ; 3 plates, 1 fig.). — The following formula is recommended as 

 highly satisfactory. The terminal buds and tender shoots were slightly injured ; 

 100 per cent, of the aphides were killed. 



Nicotine sulphate . . . . . . . \ oz. 



Whale oil soap . . . . . . . \ lb. 



Water .......... 5 gallons. 



V. G. J. 



Roses under Closk Glasses. By L. Daniel (Le Jard. vol. xxix. pp. 289,290; 

 1 fig.). — Protection can be given to delicate blooms on dwarf roses by means 

 of a clock glass supported on three oak stakes. The top and the side of the 

 glass which faces south are covered with muslin to protect the flowers from 

 scorching. — S. E. W. 



Sampling in Soil Surveys, The Probable Error of. By G. W. Robinson and 

 W. E. Lloyd (Jour. Agr. Sci. vii. Part 2, pp. 144-153 ; Sept. 1915). — The authors 

 point out that there is practically no literature on the subject. The soils of 

 two fields were therefore investigated, 25 and 9 samples respectively being taken 

 from them. In each sample the amounts of gravel, sand, silt, clay, moisture, 

 organic matter, and phosphorus pentoxide were determined. The samples 

 were then mixed and six mechanical and phosphoric oxide determinations of the 

 two composite samples were made. From these results the probable errors 



