THE PROBABLE ORIGIN 01- OENOTHEvRA LAMARCKIANA SER. 585 



"Cctte esp^cc est originaire do rAnieriquc scpttntrionale. On 

 la cultive au jardin du Museum d'Histoire Naturelle (V.S.)."') 

 The description, however, quotes some points which are not visible 

 on the herbarium specimen, nor on specimen B. it is therefore 

 clear that the author knew his plants from another source still, 

 probably from the living material of the Jardin des Plantes. The 

 most interesting point for us is the description of the fruits: "Le 

 fruit est une capsule courtc, cylindrique, glabre, tronquee I^j^'ere- 

 ment, quadrangulaire, n'ayant environ que le tiers de la longueur 

 du tube calicinal."2) This description wholly agrees with the fruits 

 of the present species, especially if we remember that Lamarck 

 based his description on a comparison with the only other large- 

 flowered form he knew, 0. longiflora. The short fruits at once 

 distinguish our species from the allied types, such as 0. suaveolens 

 Desf, and 0. grandiflora Ait., which have thin and proportionally 

 long fruits.^) 



This character of the fruits shows that the description of the 

 Encyclopedic has been based upon specimen A and not upon the 

 other one. For, although B lacks fruits also, it belongs to an 

 elementary species which has long and narrow fruits, as we shall 

 soon see. Here 1 might point out that in systematic researches 

 of this kind more value is to be attached to published diagnoses 

 and descriptions than to the material preserved in a herbarium. 

 The older systematists, as a rule, did not take much care of their 

 material, even if they were very careful of their descriptions.*) 

 The herbarium specimens are often found without their names and 

 without any indication concerning their origin. The rule "de- 



i) V. S. ("vidi siccum") means that the diagnosis is based on her- 

 barium material. 



2) Encyclopedia methodique, Botanique par Lamarck, Tome IV, 1796, 

 pp. 530—554, "Onagraire". Twelve species of this genus are enumerated, 

 O. longiflora being no. 4, O. corymbosa no. i i and O. grandijlora no. 12. 

 A copy of the diagnosis of this last one may be found in my Mutation 

 theory (p. 441) and in the article of Davis. The article in the Kncyclo- 

 pedie is not signed and was probably written by Poiret, who prepared 

 many articles in vol. IV and wrote the whole of the later volumes. In 

 the herbarium of Paris some of the specimens may be seen quoted with 

 the authority of Poiret, as, for example, on the sheet of O. suaveolens 

 Desf. , where above that name is written Oenothera grandiflora Poiret 

 Encyclopedic {Q,{. pi. 39 of the article of Davis). 



3) \J Oenothera grandiflora de I'herbier de Lamarck. O^era VI, p. 564. 



4) Cf. Bonnet, op. cit. p. 138. 



