98 THE JOUHNAL OF BOTANT 



the pen on the part of Merat, who only transfers to Leant od on a 

 Thrincia wrongly quoted by him as T. nudicaulis Lag. But no 

 such name exists. Lagasca, Gen. et Spec. Nov. p, 24 (1816), wrote 

 nudicalyx not nudicatilis, so that Herat's Leontodon must either be 

 altered to nudicalyx or treated as a nullit}^ 



II. That Leontodon hibtus of Linnaeus * was rightly inter- 

 preted by Villars and is identical with i. Villarsii Lois, results 

 from («) the characters; (h) the southern habitat ; (c) the synonyms, 

 which, although not all applicable to L. Villarsii, absolutely exclude 

 Thrincia liirta ; {d) the Linnean Herbarium, which contains two, 

 and only two, specimens named hirtiun by Linnaeus, both being 

 indisputably L. Villarsii, one received from Allioni and the other 

 grown in Hort. Ups. 



This view has been clearly stated bv Villars (1789), Wallroth 

 (1822), Kichter (1834), Koch (1887), Eischoft" (1851), and Rouy 

 (1907), and accepted by Beck (1893) and Williams (1901). But the 

 evidence from the Herbarium, which clinches the argument, is now 

 brought forward for the first time. 



The contrary theor}^, that Leontodon liivtus is Thrincia hirta 

 Roth, started with Dillenius (1719) and Buxbaum (1721), who 

 were followed by Leysser (1761), Leevs (1775), Both (1788 and 

 1797), and above all by Willdenow (1800), Avho consequently, 

 in Sp. PL iii. p. 1552, creates Apargia Villarsii for Villars's 

 L. hirtvs- — a name subsequently transferred by Loiseleur, Fl. Gall. ii. 

 p. 514 (1807), to the genus Leontodon. It is probably to the 

 influence of Dillenius, as editor of the third edition of Kay's Synopsis 

 (1724), that a plant gathered by Bobart near Oxford is there (p. 167) 

 wrongly referred to synonyms of Bauhin and Colonna, which were 

 afterwards taken up (that of Colonna onlj^ b}^ implication) by 

 Linnseus for his Crepis midicavlis and Leontodon hirtus, although 

 really belonging to L. crisjms Vill., as had been well known to Bay 

 himself. Thus were the tares sown that produced their crop in 

 the pages of Hudson, Sibthorp, Smith, Sowerby and Syme and 

 Bentham. 



To prove the identity of L- hirtvs L. with L. Villarsii con- 

 clusively, and to trace the growth of the contrary error, involves a 

 discussion of a good many pre-Linnean synonyms and other matters of 

 research which few will take the pains to seek ovit for themselves. 

 Unfortunately, it is not possible to do this concisely. As to the old 

 synonyms quoted by Linnaeus, some can be identified with cei'tainty, 

 and it is on these that we must rely ; some are doubtful ; some are 

 clearly wrong, /. e. incompatible with the others or with his diagnoses 

 and characters. The only way to make them intelligible is to take 

 the names in chronological ordei", including a few not directly refeiTed 

 to by Linnteus, These are they, and I will point out at once that one 

 only, and that doubtfvilly, can be landerstood to apply to Thrincia hirta. 



* Linnaeus wrote hiriirm, but it is surely the ne phis ultra of pedantry to 

 perpetuate the false genders of Linnaeus, or of other authors who, like him, 

 "in Greek are sadly to seek." The cases of Stachyti and Orchis are different, 

 for althovigh these words are masculine in Greek, they at any rate seem to 

 be used as feminine by Pliny, from whom they were adopted. 



