26 



PROCEEDINGS OP THE 



what plane the nut is compressed. But I iufer from the unsym- 

 metric form of the nut and the way in which the style-branches 

 invariably slope, that the nut is here laterally compressed, and 

 that the plane through the st} le-branches passes through the 

 theoretic axis. 



Fig. 5. 



Another remarkable variation of Sedges occurs (in a few genera 

 allied to Sc]icenus)hj which the style-branches become 4. I have 

 seen 4-branched styles occur, very rarely, in genera throughout 

 the Order ; but the irregularity then appears strictly accidental, 

 it occurs in only one or two flowers in the spike, and may arise 

 from an injury : it is often very imperfect, i. e. the three style- 

 branches are normally placed, one of them being bifid some way 

 up. But in the genus Tetraria, Beauvois, the style has very 

 generally 4 branches, the ovary is 4-angled, and usually there are 

 then either 4 (or 8) stamens. We may compare the analogous 

 case of Paris quadrifolia and other 4-merous Lilies. It is easy 

 to invent some possible advantage to the plant in having 4 

 (instead of 3) style-branches ; but it appears to me next to im- 

 possible to demonstrate which was the effective and causative 

 advantage, assuming that there was one. Hardly any of these 

 4-merous species have got thoroughly settled ; many of them 

 often show 3-merous flowers ; they are probably a thing of 

 yesterday as compared with Fycreus and such genera. It has 

 been maintained that "no adequate evidence is yet forthcoming 

 w^th respect to the inheritance of discontinuous variation." The 

 manner m which the 4-branched style has become prevalent ia 

 so many species of Tetraria may be put forward as evidence until 

 objectors state what further evidence is in their opinion to be 

 expected. 



