A STUDY OF BAEBAHEA VULGARIS R. BR. 205 



variety of vulgaris with the same description and mentioned as 

 rare, Loughgall, Armagh, being given as a locahty. 



Syme (Eng. Bot., ed. 3, i, 172 (1863) ) distinguishes the true 

 B. arcuata Reichb., which he regards as a subspecies of B. vulgaris, 

 by the shape of the seed which is said to be twice as long as 

 broad as compared with as long as broad in B. vulgaris. 



It will be seen from the descriptions I have quoted that hardly 

 any two authors agree as to what, beside the direction of the 

 pedicels and pods, may be considered as essential characters of B. 

 vulgaris and B. arcuata. Most British botanists after Syme seem 

 to have assumed that B. arcuata of Reichenbach was something 

 different from the plant commonly so named in this country, and 

 have rehed almost entirely on the seed character mentioned above, 

 for distinguishing it. Mr. Druce, for example (Rep. Bot. Exch. Club, 

 1906, 210), emphasises this point in a note on a plant collected by 

 Mr. Bickham from a ditch at Upton-on- Severn, Worcester, which he 

 refers to B. vulgaris var. decipiens ; he inaccurately describes the 

 seeds as broad and short, whereas on a careful examination I find 

 most of them to be at least twice as long as broad, which should 

 make it the true B. arcuata according to Syme. As a matter of fact, 

 I have found the size and shape of the seed vary to such an extent 

 in all the specimens of B. vulgaris I have examined as to make 

 this character of httle or no value for distinguishing B. arcuata. 

 Moreover, Reichenbach himself makes no mention of this seed 

 character in any of his descriptions of the plant in question. 

 Syme may have obtained this character from Reichenbach (Ico7i. 

 germ, ii, 14, 4357 (1837) — " Semina videbis eximie oblonga in 

 speciebus affinibus brevia angulata " — where they are thus figured, 

 but in Sturm the figure shows some short and some long, and I am 

 convinced that anyone who will take the trouble to look at a 

 sufficiently large series of specimens will find that not only the 

 shape and size of the seed, but the characters based on leaf-cutting, 

 density or laxity of inflorescence, size of flower, direction of young 

 and mature pods, are not permanent, but vary in almost every 

 specimen examined and sooner or later break down if we 

 attempt to correlate them. 



We are thus forced to the conclusion that B. arcuata has no 

 real existence as a species but belongs to a series of forms to 

 which the names B. arcuata, B. taurica, B. vulgaris var. clivaricata, 

 and B. vulgaris var. decijnens have at various times been applied. 

 It is interesting to note that Fernald, from whose useful paper on 

 the North American Barbareas {Bhodora, 1909, p. 134) I have 

 derived much assistance, has independently arrived at the same 

 conclusion ; he says : '' B. arcuata is often separated, at least 

 varietally, by the slightly larger flowers which are more loosely 

 arranged in anthesis, the slightly arcuate and more slender 

 siliques, and the narrower seeds ; but in the American specimens 

 examined these characters do not seem sufficiently marked to 

 make it clear that we have two different plants." " If separation 

 be maintained," he adds in a note, "the name i>'. taurica, ^hich 

 was published in the preceding year, will have to be used." 



