4. CRUCIFER^. 141 



counties, and likely enough to be found in almost all of the 

 rest, if looked for. 



89. Arabis petr^a, De C. 



Ai-ea [1***56] 7 [8 * 10] * 12 * * 15 16 17 18. 



South limit in Merionethshire and Caernarvonshire. 



North hinit in Shetland, Hebrides, Sutherland. 



Estimate of provinces 6. Estimate of coimties 12. 



Latitude 53 — 61. Highland tj'pe of distribution. 



A. A. regions. Superagrarian — Superarctic zones. 



Descends below 300 yards, in E. Highlands (Dickie). 



Ascends to 1400 yards, or upwards, in same province. 



Range of mean annual temperature 44 — 33. 



Native. Rupestral. As wiU be observed, I hesitate to 

 give this species a provincial area so wide as its reported 

 localities would indicate ; five provinces being excluded, 

 on account of their unconoborated or vaguely stated loca- 

 Uties. Hudson records it from Devonshire, which appears 

 little likely. Aiken pretends that he found it a mile from 

 Shrewsbiuy, which is still less hkely. The Flora of York- 

 shire gives a locality without any personal authority for it. 

 In Babington's Manual the species is said to occur on 

 "Alpine rocks in N. and S. Wales ;" but I do not find any 

 personal authority for South Wales. A Derbyshire loca- 

 lity, on indifferent authority, is recorded in the ' Botanist's 

 Guide.' I have not seen this species lower than 550 yards, 

 and therefore rely on Dr. Dickie, who says, " On the gra- 

 velly banks of the Dee, at BaUater, at 800 feet (42 miles 

 inland) ." It also occurs, according to the same writer, on 

 the hill of Khoil, "at 1742 feet; this last is perhaps the 

 true lower limit." 



