132 



placed in Cathaartia by Maximowicz as C. integrifolia.^ A 



decade later anotlier ot tlie three Avas tentatively referred to 

 Cathcartia bj the same author. t But in these two species the 

 dehiscence of the capsule is confined to apical chinks and does not 

 take place, as in C, villosa, by the separation of the valTCS from 

 the placental ribs throughout their length. This circumstance 

 induced Eranchet in ISSbt to refer M, integrifolia to Meconopsis 

 rather than to Catlicavtia^ and in this action Eranchet was fol- 

 lowed as regards both plants by Maximowicz in 1889. § This 

 subordination of the character derived from the stigma to that 

 afforded by the valvular dehiscence induced Tranchet in 188611 to 

 refer two species with stigmas of the original Meconopsis type to 

 the genus Cathcavlia. The expectation that in these species, M. 

 Delavnyi and 3f. lancifoliay the valves of the ripe capsules might 

 be fissile throughout w^as not fully realised, and in 1895 Franchet 

 transferred both from Cathcartia to Meconopsis.^ But Maximo- 

 wicz and Eranchet are not the onlv authors w^hose judgment has 

 been influenced by the original decision that the crucial diag- 

 nostic character in the case of these two genera is to be found in 

 the degree of separation which takes place in respect of the valves 

 of the capsule. Tlie segregation of the genera by Bentham and 

 Hooker in 1862** again depends mainly on this character, and in 

 1876, when Maximowicz was describing Cathcartia integrifolia, 

 Regol placed in Meconopsis another species in which the stigma is 

 like that of Cathcartia viUosa.'fi Since 1886 the exi^erience of 

 others has been the same; in 1889 the dehiscence character was 

 relied on by Prantl and Kundig ++, and in 1895 the &ame character 

 induced Cummins and the writer to repeat what had been done by 

 Franchet in 1886 as regards a species with a distinct style and a 

 clavate stigma. §§ In this last case, however, the action was not 

 based on the expectation, but was taken with the knowledge that 

 in the species concerned the ripe capsules have valves which may 

 separate from the placental ribs throughout their length. This 

 particular plant, C. lyrata, in Dr. Fedde\s opinion, may equally 

 well be regarded as a Meconopsis. \\\\ The experience of the past 

 tw^elve months has gone far to justify the soundness of this view. 

 The fruits of M. Forrestii, a species of the group Primulinae 

 leseribed in 1907, have now been communicated ; so have those of 

 M, Ddavayi. In both cases thev prove to he long and cylindric 

 like those of the species which Eedde has renamed M. lyrat^- 

 They are larger than the capsules of M, lyrata, and those of 31. 



* Bull. Acad. Petersb. vol. xxiii. p. 310. 



t Flora Tjmgntica, t. 23, on which the name Cathcartia ptmicea is 

 inscribed. 



I Bull. See. Bot. Pr. vol. xxxiii. p. 389. 

 ^ Flora Tanerutica, text at pp. 34, 35. 



II Bull. Soc. Bot. Fr. vol, xxxviii. pp, 390, 39L 

 f Journal of the Asiatic Society of Bengal, vol. Ixiv. part 2. p. Hll. 



( 



tt 



inexact, in that it shows the stiema as subtended by a style 

 tt Die Naturlichen Pflanzenfamilien, vol. iii. part 2, p. 141 

 §§ Journal of the Asiatic Societv of Benr^l, vol Ixiv 

 III! Da^ Pftaiizenreich, 40 Heft [IV. 104"^rp/216 



; tlie figuiv given is 



