ANAGALLIS ARVENSIS AND A. CHRULEA 869 
Variations i ri the colour of the flowers of arvensis are recorded, 
a pink eye (Grove edition of Babington’s Manual, 1904, p. 348) ; 
purple, green ec or tinged with purple (Pryor, Flora of Hert- 
stones ‘e, 1887, p. 842) ; very pale lilac, var. pallida (Purchas & Ley, 
Flora of Herefordshire, 1889, p. ae dull blue, and blue. The 
corolla-segments in cerulea are no t really narrowed to the apex, but 
in consequence of the sides towards the apex being inflexed ra 
have that appearance. This circumstance, together with the com- 
In the o wers of arvensis, owing to the siete na shorter 
length of the c ealyx-segments, and the flat position of the corolla- 
lobes, — the extreme apices of the former are visible from above. 
On ination with a hand-lens of the hairs on the edge of the 
corolla i in n the two species, it appeared to me that those “et c@i int 
e 
rmed by examination under the microscope; from which it is 
clear that in arvensis these hairs are normally composed of three 
cells, of which the ultimate is large and clavate, whilst in cerulea 
the hairs are normally composed of four cells, of which the ulti- 
mate is large and oblong, with the sides usually feebly excavate. 
Smith, writing in 1798, had not then found any specific differ- 
ence between the red and blue pimpernel. The plate of the latter, 
published by Sowerby, Dec. ist, 1807 (E. B. 1823), gives a recog- 
nizable figure of the whole plant, but the outline enlarged drawing 
of a single flower is very inexact, as it represents the edge of the 
corolla jagged throughout, whereas its lobes are, in fact, denticulate 
only near the apex. The letterpress to this plate does not serve 
to elucidate the distinctions between the two plants, but it is pie 
of note that the obvious ciliation of the corolla in arvensis di 
escape the attention of a artist (KE. z 529), whilst he failed to 
record any hairs whatever on the edge of the corolla of cerulea. In 
Hooker's British Flora (fifth edition, 1842) the corolla of arvensis is 
described as having the margin crenate, piloso-gla ndulose, and that 
i scarcely at all glan 
further on, however, we read: ‘‘ The Rev. Professor Henslow he 
our two poo are distinct species, ‘but ‘that each varies with 
same tints of colour.” A somewhat more intelligible jand eo of 
