CllEPlS NUD1CAULI8 L. AND LEONTODON HIRTUS L. 

 Bv C. C. Lacaita, F.L.S. 

 1> the following paper I shall attempt to prove : 



I. That Crepis nudicaulis L. is either Leonfodou Villarsii 

 Lois., in which case the name must be altered to Lf;oittudoii 

 Hudicauiis, or a nomen confusum to be discarded. 



II. That Leonfodon hirlns L. is L. Villarsil and not 

 Thrincia liirta Koth, introducing evidence from the Linnean 

 Herbarium that has hitherto been overlooked. 



III. That the legitimate title for Thrincia liirta, lioth is 

 Thrincia taraxacoides (Vil!. sul) Uyoseride). 



I. Ckepis iSL'UfCArLis L. Sp. PI. p. 8Uo (1758), " foliis lanceo- 

 latis, dentato-sinuatis, hispidis, setis subulatis, scapo unilioro," is 

 certainly not Thrincia hirfa Iloih, but either Leoniodon V illarsii 

 Lois, or a mixture of that species with L. crisj)ics Yill. The Bauhin 

 synonyms cited belong to X. cri^^ij^ns or to the almost identical 

 L. saxafilis Rchb. The words " foL lane, dentato-sinuatis" are 

 more suitable to crispns than to Villamii, but the expression "setis 

 subulatis '" admirably indicates a peculiarity of Villarsii and is 

 inapplicable to any other species that could be intended. The 

 habitat, " Gallia Narbonensis et Hispania," suits both ci-ii>-j)us and 

 Villarsii, bvit hardh' would allow of Thrincia hirla. The seta^ on 

 the leaf of L. criajms are slender, but very conspicuously three- 

 forked at the top, as can easily be seen by the naked eye. Those of 

 Thrincia hirta are much less copious, short, and for the most part 

 obviously forked — " bitides en y-grec," in Villars's words. Thote ef 

 Villarsii were described by Ball in Ann. Nat. Hist. vi. (1850) as 

 " longis albis simplicibus, interdum apice brevissime f urcatis." They 

 are very thick at the base, altogether longer and stronger than in 

 crispiis or in T. hirta, but to the naked eye they look simple {cf. 

 Yill. Dauph. iii. tab. xxv.), though a good lens I'e veals tliat man}" are 

 very shortly two- or occasionally even three-pronged at the tijj. 



In 8p. PI. ed. 2, p. 1134, Linmeus quotes his previous diagnosis 

 of Crcpis nudicaulis, without mentioning that name, as a sjnonym 

 of Leonfodon hirtus, which, as will be shown hereafter, is unques- 

 tionably identical with X. Villarsii Lois. But in 8yst. Nat. ed. x. 

 p. 1194, whei-e he first created L. hirtus, Crepis nudicaulis is not 

 referred to that species, but hesitatingly thought to be a variety of 

 L. hispidus. There is no specimen of Crepis midicaulis in the 

 Linnean Herbarium. There might therefore be some excuse for 

 rejecting the specific name nudicaulis as uncertain and ambiguous, 

 but if used at all it must, as the oldest, be substituted for Leontodon 

 hirtus L. and consequently for L. Villarsii Lois. It cannot be sub- 

 stituted for Thrincia hirta lioth (whether that species be transferred 

 to Leontodon or not), as suggested by Mr. Britten in Journ. Bot. 

 xlv. p. 81 (1907). Leontodon midicaulis Mc'rat in Ann. Sci. Nat. 

 xxii. p. 109 (1S81), quoted by Williams, Prodr. Fh Brit. i. p. 70 

 (1901), as synonymous with T. hirta, is a mere misprint or sliji of 

 JoLU^AL UJ' IjOTANY. YoL. 50. [Al'KJL. 1918.]- n 



