1S3 



Forrestii liaie liaidly any style. But size is not an important 

 feature, Avliile tlie difference as regards style, as the cases of 

 il/. integrifolia and A/, fseudointegnfolia, and again, of M. 

 chelidoni folia and M. Oliveriana liave shown us, need not be of 

 more than specific importance.* JVIoreover, two new members of 



group 



d cylindric, but have well-developed styles lite that 

 Among the &^(}cies of Meconopsis with long cylindric 



case 



by the short apical valves of a typical Meconopsis or a true Papa- 

 ver. But in M. Forrestii, M. Delavayi^ M. venusta, and 3L 

 conciiina, the extent to which the valves separate is considerc^bly 

 greater^ and these four forms thus supply conditions intermediate 

 between those that occur in i¥. Oliveriana and in M. lyrata. The 

 dehiscence character, therefore, turns out to be a relative,^ not an 

 absolute one ; while it may still prove of value for specific dis- 



remains 



J 



allied genera. 



When tlie dehiscence character was belieTed to be of g 

 significance-— and nntil direct evidence that it had broken down 



the treatment 



was available there was no good reason for questioning this belief 



■ ' '" a 3^econopsis and Cathcartta 



in the Pflanzenreicli was inevitable, ^'"ow that the inadequacy of 

 this character is manifest the general position must be reviewed, 

 and the first consequence of the altered situation must be the 

 transfer of the section Cumminsia from Cathcartia to Meconopsts. 



Lsider whether the sec- 



^»,^^^..^^...^ ^. - ^ ^o Meconopsis. In 



arriving at a conclusion, it is necessary to_ estimate the value of 

 the character afforded by the Cathcartia stigma. _ ^ ; 



We have seen that this character was accorded generic sigmii- 

 cance bv Masimowicz in 1876, and the fact that after 1886 



tran 



Ma 



does 



was erroneous. Tranchet's judgment of 1886 was induced by the 

 fact that when ripe fruit's of Cathcartia integnfolia became 

 known thev were found to onen only by apical valves. It was 

 therefore based upon the earlier iudgment of Hooker which in 

 1886, there was no reason to doubt. Now that it is known fhat 

 the dehiscence character is, at best, of no greater value than the 

 stigmatic character, it is permissible to enquire whether the latter 

 criterion can be adopted in substitution of the former.^ The recog- 

 nition of a genus Cathcartia, enlarged as Maximo wicz m eftect, 

 suggested in 18T6 that it should be, would involve tlie transfer to 

 the original Cathcartia of the group Grandes, m which the stigma 

 agrees with that of Cathcartia, as a compensation fo/ the removal 

 of the section Cumminsia, which agrees with Cathcartia^ as 

 regards dehiscence. This, however, we know now to be im- 



pgeudointeqrifi 



- 1T> the case or ^'\^7;«y:V;7": ^'^^ ;;.- i^ fcl,^ Gardmers' ChronicU iu 

 lOoVrand las^ r^^YS^n ^^ to a disposition in so.ne 



quartf^s to Sr^Leter the presence or absence of a style be even of 



Hpeci 



