lience no violent assumption to make, that as D. oleoides 

 extends from Spain to the Himalaya, so may caucasica 

 extend from the Caucasus to the Altai. If this be so, the 

 choice of names becomes a question, for both were published 

 in the same book by Pallas. Of the two D. altaica is the 

 better known, the broad-leaved form of it having been 

 figured in this magazine (t. 1875), from a plant that 

 flowered in the Cambridge Botanical Garden in 1817. 

 The latter are described as not being fragrant, whereas 

 those of D. caucasica are decidedly so. On the other hand, 

 Pallas's plate of D. altaica is quite erroneous, and contrary 

 to his description, in figuring the clusters of flowers as 

 supported by a long peduncle. 



Plants of D. caucasica were procured by the Royal 

 Gardens in 1893, under the name of salicifolia, from Mr. 

 T. Smith of Eewry, whose nursery is famous for the 

 number of rare shrubs it contains, many of them, though 

 hardy in Ireland, unfortunately not so at Kew. In the 

 case of D. caucasica, however, it stood the winter of 

 1893-4, and flowered in May of the latter year, but has not 

 fruited. 



Descr. — A dwarf shrub, quite glabrous, except the 

 perianth. Leaves one to one and a half inches long, 

 deciduous, linear-lanceolate or oblanceolate, subacute or 

 obtuse and apiculate, pale green above, subglaucous 

 beneath. Flowers subsessile, in terminal clusters of two to 

 twenty, white, fragrant, ebracteate. Perianth tube half an 

 inch long, cylindric, silkily pubescent ; lobes ovate or nearly 

 orbicular, about half as long as the tube, margins at first 

 involute, then revolute. Stamens included, except the 

 tips of the four upper anthers. Ovary obovoid, sparsely 

 pubescent ; style very short ; stigma broad, hemispheric. 



Pig. 1 Flower; 2, the same, with the perianth laid open; 3, ovary; 

 % vertical section of the same :— All enlarged. 



