182 7. VIOLACE^. 



largest-flowered specimen of the Scottish V. lutea, purple 

 variety, which I have ever seen, came to me from the same 

 Society, mider name of " Cm-tisii," though located fi-om 

 " Glen Tmi-it," Perthshiie. Doubtless, the botanists who 

 manage that Society know better, and this odd crossing of 

 names must have arisen from occasionally employing la- 

 bel-writers who were not sufficiently carefiil in taking down 

 names. 



137,c. Viola Curtisii, Forst. -^ hf •///.( ^^^, 



Ai-ea 1 ^ -;fj ^ j^ 6 7 -;;. 9 * * * * * [15]. 



Soutli hmit in Cornwall or Devon. 



North Umit in Anglesea or Cheshire. 



Estimate of provinces 4. Estimate of counties 6'. 



Latitude 50 — 54. Local ij^Q of distribution. 



Agrarian region. Inferagiarian zone. 



Descends to the coast level, in the Peninsula. 



Ascends, on the coast level, to the Mersey } 



Range of mean annual temperatm-e 52 — 49. 



Native. Littoral and Glareal. I treat the distribution 

 of this peculiar plant distinctly, not so much from suppos- 

 ing it to be a permanent species, as from inability to 

 form any satisfactory conclusion whether it is nearer to tri- 

 color or to lutea. In the last edition of the British Flora 

 it is slightly mentioned as a yellow variety of V. tricolor ; 

 while in the Manual of British Botany it is placed as a va- 

 riety of V. lutea. As a geographical botanist, I should 

 find convenience in sinking Cmtisii imder tricolor ; the 

 area and cHmatic range of the latter, V. tricolor, compre- 

 hending those of Curtisii, which is not the case with the 

 ai'ea and range of V. lutea. But against convenience I am 

 compelled to set up the obstacle of V. Curtisii appearing 

 nearer to V. lutea in its technical characters, cs]iecially in 



