84 ORIGIN OF CULTIVATED PLANTS. 



genous may in several cases be the result of self-sowing 

 from cultivation,^ or that the species was formerly com- 

 mon, and is tending to disappear. Its presence in the 

 western islands of Lurope favours the latter hypothesis, 

 but its absence in the islands of the Mediterranean is 

 opposed to it.^ 



Let us see whether historical and philological data 

 add anything to the facts of geographical botany. 



In the first place, it is in Europe that the countless 

 varieties of cabbage have been formed,^ principally since 

 the days of the ancient Greeks. Theophrastus dis- 

 tinguished three, Pliny double that number, Tournefort 

 twenty, De Candollo more than thirty. These modifica- 

 tions did not come from the East — another sign of an 

 ancient cultivation in Europe and of a European origin. 



The common names are also numerous in European 

 languages, and rare or modern in those of Asia. Without 

 repeating a number of names I have given elsewhere,* I 

 shall mention the five or six distinct and ancient roots 

 from which the European names are derived. 



Kap or hah in several Keltic and Slav names. The 

 French name cahua comes from it. Its origin is clearly 

 the same as that of caput, because of the head-shaped 

 form of the cabbage. 



Caul, Icohl, in several Latin (cavMs, stem or cabbage), 

 German (Ghdli in Old German, Kohl in modern German, 

 Icaal in Danish), and Keltic languages [kaol and kol in 

 Breton, cal in Irish).^ 



Bresic, bresych, brassio, of the Keltic and Latin 

 (brassica) languages, whence, probably, be7'za and verza of 

 the Spaniards and Portuguese, varza of the Roumanians.^ 



' Watson, who is oaroful on these points, doubts whether the cabbago 

 is indigenous in Eufjlainl {Oompendium of the Cyhele, p. 103), but most 

 authors of British floraa iidmit it to bo so. 



' Br. balearica and Br. cretica are perennial, almost woody, not 

 biennial ; and botaniats are agreed in separating thorn from Br. olerarea. 



' Aug. I'yr. do UandoUo has published a paper on the divisions and 

 subdivisions of Br. oleraciu [Transactions of the Hart. Soc, vol. v., tran!=- 

 lated inl^o German nn 1 in French in the BM. Univ. Agric, vol. viii.), 

 whioh is often quoted. 



* Alph. de Candollo, Qingr. Bat, RaisonniSe, p. 889. 



' Ad. Pictet, Les.Origines Indo-Kumpiennea, edit. 2, vol. i. p. 380 



• Brandza, Prodr. Fl. Komane, p. 122. 



