102 THE .TOU]l?iA.L OF BOTANY 



and the additional characters ; " Foha pilis apice indivisis (2) ; radius 

 corollffi subtus etiara luteus nee vireseens (3) ; pappus plumosus (4)." 

 The diagnosis, but not the name, of Crepis nudicaulis is now quoted 

 as a synonym, as well as L. hispidum vars. /3. and y. The habitat is 

 Helvetia, Grallia Narbonensis, Hispania. 



(C) In Syst. Nat. ed. xii. p. 522 (1767), with the same diagnosis 

 and the further characters : " Folia rigidula, tactu quasi arida (5), 

 laciniis obliquatis (6) ; cal3"cis foliola oxtima dilatata mediante 

 plica." No synonyms or habitat are mentioned. 



(D) In Mantissa, p. 458 (1771J, no new characters are assigned; 

 four fresh synonyms are introduced ; Crepis nudicaulis is not alluded 

 to, but L. hispidum vars. /3. and y. are excluded and referred back to 

 their original position under that species, apparently in deference to 

 the arrangement in Gerard, Prov. p. 166. 



Of the various characters no less than six — those numbered — are 

 incompatible with Thrincia hirta or with any other species than 

 L. Villarsii: (1) " setis simplicissimis " and (2) "folia pilis apice 

 indivisis " are different ways of expressing that without considerable 

 magnitieatiou the hairs appear simple, as shown in Villars' figure of 

 his L. hirfus, Dauph. iii. tab. xxv. But in Thrincia hirta, as in 

 L. crispus, they visibly fork into two or three points at the tip. 



(3) " Kadius corollaB subtus etiam luteus nee vireseens." This 

 excludes both L. hispidas and Thrincia hirta, in both of which the 

 difference of colour in the lower surface of the ligules, varying from 

 greenish to livid or reddish brown, is conspicuous and still plainly 

 visible in old exsiceata of the middle of last century. But it does 

 not exclude L. crispus, which also has ligiiles yellow underneath. 



(4) '"'■Pappus plumosus'''' : this seems to exclude any Thrincia. 



(5) and (6) " Folia rigidula, tactu quasi arida " and " laciniis obli- 

 quatis." These leaf characters are very distinctive of L. Villarsii 

 and quite fatal to the claim of Thrincia hirta. Leontodon crispus, 

 if not so obviously excluded by (5), is quite incompatible with (6), a 

 phrase which may be compared with Ray's "folia divisione Erucse 

 purpurese." Thus the character are even more decisive than the 

 synonyms in favoiar of L. Villarsii. 



Now for the Linnean specmiens. There are four sheets pinned 

 together *. But of these only two have been named by Linnpevis 

 himself, and both are precisely L. Villarsii. Another is Leontodon 

 pyrenaicus Gouan, and the fourth is Thrincia hirta Roth. But 

 neither of the latter have been named by Linna?us, or indeed bear any 

 writing of his except the abbreviation "All." on the sheet oi L. pyre- 

 naicus. Three of these specimens, viz. the two L. Villarsii as well 

 as the Thrincia hirta bear the words " hirtum D. Afzel." written in 

 pencil by Smith. How Smith obtained Afzelius's determination of 

 them does not appear ; but it is worthless, confusing as it does two 

 such very dissimilar plants. They have also been examined and 

 ticketed by Schultz Bipontinus : unfortunately his ticket, now 

 attached to the uppermost and most important specimen, has obviously 



* Dr. Daydon Jackson informs me that he has not been able, in the course 

 of his researches into the history of the Linnean herbarium, to ascertain to 

 whom the pinning together of sheets is to be attributed, though in some cases it 

 may liave been done l>y Linnajus himself. 



