ROSACE.E. 2 15 



coarsely serrated ; veins 5 to 8 on each side, prominent beneath, 

 Calyx-sec^ments spreading in flower, erect in fruit. 



On exposed rocks, partieuhuly those of limestone or trap. 

 Sparingly but generally distributed. I have specimens from the 

 counties of Devon, Derby (Matlock Bath), Carnarvon (Ormes Head), 

 Durham (Holy Island), Yorkshire (Teesdale), Edinburgh (Quezon's 

 Park), Sutherland, and from Sligo (Ben Bulben), Antrim (Gleiiarm). 



England, Scotland, Ireland. Shrub or Tree. Early Summer. 



Usually a bush 3 to 10 feet high, closely resembling P. eu- 

 Aria, but with the leaves wedge-shaped towards the base and 

 broadest beyond the middle, less deeeply lobed, and with a 

 greater portion at the base entire than in the P. eu-Aria; the 

 upper surface becomes glabrous much sooner ; the veins being 

 more distant give the leaf a less ribbed appearance, as both in this 

 and the preceding plant the veins are depressed on the upper and 

 prominent on the lower surface. The partial corymbs which com- 

 pose the general corymb are denser, the flowers a little larger, the 

 calyx-segments not reflexed and rather narrower than in P. eu-Aria. 

 The mature fruit I have never seen. 



I have had much hesitation in separating this from P. eu-Aria, 

 but there appeared to me no possible medium between adopting this 

 course or making P. Aria, P. scandica, and P. fennicamcre vai'ieties. 

 All the four forms pass insensibly into each other, and yet the 

 extreme forms of each are too far ajiart to suppose it possible that 

 the one could be derived in a limited time from the other. I am 

 indebted for fresh specimens of the Derbysbire plant, from which 

 the accompanying plate has been made, to Mr. Joseph Wiiittaker, 

 for the Irish to Professor Dickie, for the Edinburgh to Mr. John 

 Sadler. The Teesdale specimen is from Mr. Baker, who kindly 

 sent me leaves from the diiferent forms of P. Aria, of which he 

 possessed specimens. 



I regret that I cannot retain either Reichenbach's or Fries' 

 specific name, in consequence of not separating the genera Pyrus 

 and Sorbus : there is already a Pyrus oblongifolia of Spach and a 

 Pyrus salicifolia of Linnanis. 



Rock White-beam. 

 Sub-Species III.— Pyrus scandica. Bah. 



Plate CCCCLXXXIV. 



Bah. in Bot. Gaz. Vol. III. p. 35 ; and Man. Brit. Bot. ed. v. p. 118. 



Sorbiis scandica, Fries. Nov. Fl. Suec. p. 138; Sum. Veg. Scaud. pp. 42, 175. 



Cratsegus Aria, var. o, scandica, Linn. Amcen. Ac. Vol. II. p. I'JO. 



C. Aria, p Suecica, Linn. Sp. Plant, p. G81. 



Pyrus intermedia, " Ehrh." Lindley, Syn. Brit. Fl. p. 105. 



P. aria, var. y. Hook. & Am. Brit. Fl. ed. viiL p. 141. 



