BRITISH ELPHRASi.E 21 



large-flowered form must be regarded as the true E. campestris 

 Jordan, which is a very well-marked plant, not yet found in Britain ; 

 that this is well separated by definite characters from the small- 

 flowered form, and that the*^ latter should be distinguished by a 

 varietal name. I therefore ]jropose the name of var. neglecta, not 

 disregarding the possibility that further observation and study may 

 show that some of the British plants are of hybrid origin. 



The following scheme will show concisely the characters which 

 separate E. campeafrls, the var. neglecta, E. Rostkoviana and 

 E. hrei'ipila : — 



Corolla large, its tube elongating at the end 



of the flowering 1 . 



Corolla smallpr, its tube not elongating 2. 



1. Stem branched below, internodes long, spike 



lax except at the top, leaves and bracts 



large, glandular hairs long E. Rostkoviana. 



— Stem branched above, often corymbosel}^ 



branches crowded, spike dense, leaves and 



bracts smaller, glandular hairs shorter ... E. campestris. 



2. Stem branched below, internodes long, spike 



broad with long internodes below, leaves 

 and bracts broadly ovate with numerous 

 teeth E. hrevipila. 



— Stem more often with more spreading branches 



above and below the middle, spike more 



slender, leaves and bmcts more narrowly [var. neglecta. 



ovate with fewer more acute teeth E. campestris, 



England. — ]S'. Somerset. Failand near Bristol, Sept. 1898, 

 C. Bucknall Eecorded in White's Bristol Flora (1912) as 

 E. hrevipila. In 1916 this was observed from July to the end 

 i)i September, and its known area greatly extended. It is plentifully 

 distributed at intervals along the limestone ridge on grassy roadsides 

 and in rough pastures from Failand to Cadbury Camp, a distance of 

 1 miles, and is everywhere mixed with E. nemorosn. To this it is so 

 similar in habit that I have only been able to distinguish it by the short 

 glandular hairs. I have suggested above the possibility of its being 

 of hybrid origin, and this point requires further investigation. — In a 

 thicket, Charlecombe Bay near Clevedon, Sept. 1916, /. M. Roper-. 

 fine specimens attaining 10 cm. in height. — Glamorgan. Oyster- 

 mouth near Mumbles Head, Sept. 1909, C. Bailey. Stem nearly 

 30 cm. high ; similar to the Somerset plants. — I)erry. Matlock, 

 Aug. 1915, and Middleton-by-Wirksworth, Sept. ]91o. These, kindly 

 sent by Dr. Drabble, differ slightly in the branches springing from 

 the stem at a sharper angle, in the principal leaves being less 

 spreading, and in the rather broader corolla. 



13. E. GRACILIS Fries. Stem slender, wiry, dark purple, simple 

 or branched about the middle, internodes often very hjng, seldom 

 short ; branches much shorter than the stem, erect. vSpike long and 

 slender, lower internodes long. Leaves 5-7 mm. long. Bmcts 

 5-6 mm. long, the upper with lanceolate or subulate acute oi,"^ 



