il2 THE JOURNAL OF BOTANY 



say, *S'. lancifolia, anruiensis, subpinnata, siibsimilis, seem to have 

 originated in Norway. Then seeds of S. arranensis (the form 

 with small and whitish anthers, which is well-distributed in 

 Hardanger), may have been carried by birds in winter-time to 

 the Scottish island of Arran. That this was the case is confirmed 

 ■by the fact that S.feimica, which is widely distributed in Norway, 

 likewise occurs in the same island. It should be added that 

 Sorhus intermedia, which occurs so frequently in Smolandia 

 (Sweden), especially eastwards, and (according to E. Fries) forms 

 many small w^oods there, that it may be assumed to have 

 originated here, and spread hence, has lately been detected in 

 Scotland, according to specimens kindly communicated by Eev. 

 A. Ley. In 1909 six trees of this species were found by Eev. 

 E. S. Marshall and Dr. Shoolbred near a stream at Garve, 

 E. Eoss-shire, Scotland [see Journ. Bot., 1910, p. 134, E. S. M.] ; 

 and they were jointly of opinion that the biggest tree was a 

 hundred years old. As S. intermedia has only been cultivated in 

 England and Scotland lately [temporibus ultimis] , it is easily 

 understood that these trees grew from seeds brought by birds 

 from Scandinavia." 



It has been asserted that birds normally migrate /as ^^7^(/; and, 

 although Sorhus-hmt^ may take a considerable time to digest, and 

 the seeds may be retained still longer, before being voided, yet 

 the British distribution of S. arranensis and S. fennica (only in 

 ihe luestern Isle of Arran) presents a further difficulty. 



According to Nyman, Cratcegus scandica L. am. ii. is a 

 synonym of S. scandica Fr. ; and this may make the name 

 S. intermedia Pers. (Pijriis intermedia Ehrh.) invalid. 



II. SoRBUS NORVEGiCA Hedl. nov. nomen. Syn. Fyrits Aria 

 a oht'usifolia De Candolle, Prodr. ii. p. 636, pr. p. (Fl. Dan. 

 302, pr. p. S. rupicolam respicit) 1825. Sorhus (Aria) 

 ohtusifolia Hedl., Monogr. d. Gattung Sorhus (Kongl. Sv. 

 Vet. Akad. : s. Handl. Bd. 35, N : o i, p. 80), 1901, nomen 

 jam antea variis Sorbis attributum. Eeprinted from Nyt 

 Magazin for Naturvidenskahene, Band 52, H. 3 [1914] , 

 pp. 255-9. 

 S. norvegica Hedl. is fully described. " Its distribution is 

 mainly in Southern Norway, more rarely in Bahusia (Sweden), 

 and it was formerly reported from Gotland and Lilla Karlso : 

 apparently it is absent outside Scandinavia. Differs from the 

 subsimilar forms of S. Aria, especially by the irregularly developed 

 .pollen, as is the case in the heterozygotic plant S. Auc^qjaria x 

 S. Aria. It should be most closely compared with S. porrigens, 

 n. sp., not yet described, which is distributed through Central 

 Europe from Asia to England. S. porrigens differs from S. nor- 

 vegica in the smaller leaves, shortly obovate, more decidedly 

 <}uspidate, with the serratures of the upper third of the leaf very 

 ■often strongly porrect. I have seen specimens from Asia Minor : 

 Paphlagonia (P. Sintensis, Iter Orientale, 1892, No. 5128, pr. p. ; 

 the specimens belong in part to S. grceca) ; Balkan : Akdagh 



