348 ADDITIONAL SPECIES, ETC. 



Arran; so that I feel compelled to join it with P. pimia- 

 tifida, although certainly an approximation to P. Aria by 

 its less deeply pinnatifid and more do^vny leaves. With 

 those examples before me, I cannot limit the area of P. 

 pinnatifida to the Isle of Arran, as Mr. Babington would 

 do (Manual, page ill; and Bot. Gaz. vol. iii. page 35). 

 From the localities reported in Somerset (Mr. Flower) and 

 Surrey or Hants (Mr. Eeeves ; Bromf. Cat.) I have seen 

 no specimens. (See vol. i. page 367.) 



365,0. PyRUS INTEEMEDIA, jE/tr/j. ^L€. /* ■ ^~^ / - 



(Pyrus scandica, Bah. ed. 3.) 



Area 1 a 3 * * * 7 * ,^ * 11. 



South limit in Devon, Hants, ? 



North limit in Durham, Denbigh, ? 



Estimate of provinces 6. Estimate of counties 12. 



Latitude 50 — 55. English tjqae of distribution. 



Agrarian region. Inferagrarian — Midagrarian zones. 



Descends to the coast level, in the Peninsula. 



Ascends to 300 yards, or upwards. 



Eange of mean annual temperature 50 — 45. 



Native. Eupestral, Sylvestral. I have seen this alleged 

 species in Devon and Denbigh. To these counties Mr. 

 Babington adds (Bot. Gaz. iii. p. 35) Somerset, Hants, 

 Berks, and Durham ; and Mr. Lees reports that of Caer- 

 narvon likewise. Difficult to be distinguished from P. 

 Aria, in the case of many examples which appear like 

 transition forms. (See vol. i. page 367.) 



