345 



those species referable to the genus Plocandra as understood by 

 Meyer that do not conform in character with the definition of 

 Pseudosabbatia. At the same time Boeslinia, Gilg, is more unlike 

 the natural group Boeslinia, Endl., which corresponds to Moench's 

 original genus so named, than the composite section Boeslinia, 

 Griseb. Gilg, following Grisebach, has retained in Boeslinia the 

 species C. serpyllifolia, which is a Hippochiron ; he has further 

 admitted to the section a species from Madagascar which belongs 

 to another subgenus. Inadvertently Gilg has attributed to all the 

 species included in the section a berry-like fruit. This is not the 

 case as regards C. serpyllifolia, the fruit of which is a capsule, as 

 Grisebach has described it. Adopting an accidental modification 

 in its orthography used by Bentham and Hooker, Gilg has 

 followed Grisebach in doubtfully reducing Valerandia, Neck., 

 to Orphium, E. Mey. 







Valerandia was proposed by Necker (Elem., ii., p. 33), to include 

 some of the species of Chironia, Linn. The diagnostic character 

 of the genus is the existence in the flower of a ring of crenate 

 Bca ! et ?' . These scales, according to the sequence of characters in the 

 definition, lie within the corolla. The one reliable diagnostic 

 character of Orphium, E. Mey., is the existence of a crenate disk, 

 not met with in Chironia. This disk lies outside the corolla. 

 Therefore the existence of this disk, so far from suggesting that 

 Orphium, E. Mey., may be the same as Valerandia, Neck., shows 

 that Valerandia cannot be Orphium. The evidence to this effect 

 does not end here. Necker has given a definition of Chironia 

 {Elem., ii., p. 34), as limited by him after (a) the segregation of 

 whatever Linnean species he had placed in Valerandia, and (&) the 

 exclusion of C. baccifera {Boeslinia, Moench). The diagnostic 

 character of this restricted genus Chironia, Neck., is that the 

 anthers are spirally twisted. Therefore Chironia, Neck., does not 

 include any true Chironia enumerated by Linneaus, but is 

 restricted to the three Linnean Chironiae which belong to the 

 genus Sabbatia, Adans., and to C. frutescens which is Orphium, 

 E. Mey. Valerandia then, as described by Necker, cannot be 

 Orphium, E. Mey. ; Chironia, Neck., after the segregation of 

 Valerandia, still included C. frutescens, Linn., the plant on which 

 Orphium, E. Mey., is based. 



The eflfect of Necker's diagnosis of Chironia is reflected in the 

 description by Gaertner in 1791 ■ {Fruct., ii., p. 156), in which stress 

 is laid on the character of twisted anthers. We know that Chironia, 

 Gaertn., was less circumscribed than Chironia, Neck., for Gaertner 



id partially figured the species C. baccifera whicn 

 expressly excluded. The character of twisted anthers 



included 



^ used by Necker, was probably quite diagnostic ; as employed by 



baccifera 



that Valerandia, 



it made bv Grise- 



bach in 1845. There is no ring of crenate scales withm the corolla 

 f any of the species belonging to the genus Chironia as hmited 

 b J Endlicher himself in 1838. What Valerandia, Neck., may be 

 foes not concern the present enquiry ; it is suflicient here to show 



Jhat Valerandia, Neck., cannot be Orphium, and cannot be 



wironia. 



