272 THE JOUK>'AL OF UOTANY 



them separately in the Gardeners Dictionari/ in 1768, but applied 

 the Linnsean name to the Bermuda plant and renamed the Atlantic 

 coast species as *S'. angustifoliuin. William Curtis, who, like Miller, 

 knew both plants, also considered them to be distinct, and in the 

 Botanical Magazine, plate 94, named the Bermuda plant S. iridioides ; 

 the date of the titlepage o£ volume 3 of the Botanical Magazine is 

 1790, but the printed date on the plate itself is September 1, 1789 ; 

 the publication of the binomial must, therefore, date from that of the 

 plate, 1789. 



Modern botanists follow the interpretation of Philip Miller by 

 applying the name Sisyrinchium Bermudiana L. to the plant that is 

 endemic in the Bermudas, but this is contrary to the laws of priority 

 as expressed in both the Vienna and American codes. Both of these 

 species were described and illustrated by Flukenet in the Almagestum 

 [p. 348, t. 61] under his genus Sisyrinchium; likewise by Dillenius 

 in Hortus Elthamensis under the Tournefortian genus Bermudiana. 

 Linnseus, in the Sjjecies Plantariim, p. 954, 1753, combined both 

 species under the binomial Sisyrinchium Bermudiana, thus preserving 

 to science both of the old generic names under each of wliich the 

 species had previously been known. The specific name Bermudiana 

 perpetuates an old generic name, and cannot be considered as having 

 been given to the species as a geographical name to indicate the 

 nativity of the species ; had that been the idea actuating Linnseus, he 

 in all probability would have given it the name bcrmudiense, adopting 

 it from Plukenet, providing he had intended the Bermudian plant to 

 be the type of the species. But Hemsley has already shown {Journal 

 of Botany, xxii. 108-110, 1884) that Linnseus in all probability had 

 never seen the plant from Bermuda. As a matter of fact he made 

 the Bermuda plant his var. /3, and considered it to be of such small 

 categorical importance that he did not give to it even a varietal 

 designation. That he intended the Virginia plant to represent typi- 

 cally his S. Bermudiana is clearly proved by the fact that all references 

 to it were enumerated under his specific name and description, while 

 those referring to the Bermuda plant were grouped under his unnamed 

 variety /3 ; and by the fact, which is still more to the point, that the 

 explanatory note with its fuller description was drawn entirely from 

 his " Planta a," /. e., the Virgmia plant. A careful study of all the 

 evidence seems to indicate that : 



1. Linnseus probably never saw the plant from Bermuda. 



2. The specific name Bermudiana perpetuates an old generic name, 

 and was not used as a geographical name to indicate the origin of the 

 species ; this view per se would prevent the adoption of the Bermuda 

 plant as the type of the species. 



3. The Linnsean descriptions (diagnosis and footnote) are based 

 upon the plant from Virginia, which must therefore be taken to be 

 the type of the species. 



4. The plant from Bermuda should be known under the first name 

 applicable to it, *S'. iridioides Curtis. 



