THE PHANEROGAMS OF THE JUAN FERNANDEZ ISLANDS 



'39 



hardly a reliable character. The standard is more orbicular in the latter, if 

 always, I cannot tell, and the flower smaller. Philippi tried to find some 

 difference in the calyx, but without much success (comp. my figures 12 d, e and g). 

 The seeds of the Chilean plant are ellipsoidal, slightly compressed from the 

 sides, castaneous, 6 — 7 mm long and 4 — 4,5 mm broad. I have not seen the 

 seeds of 6". grandiflora ; in microphylla I found them dirty yellow or yellowish 

 brown, see plate 20, fig. 3 — 4. For comparison also the seeds of S. toromiro 

 (Phil.) Skottsb. were figured, differing from those of all the others in colour, 

 size and shape. 



Fig. 11. a— c Sophora fernandeziana: a leaf of no. 63, lower side, b of no. 214 (f. gracilior), 

 nat. size; c mature pod of no. 599, £ nat. size, d— g .!>'. masafuerana :■ d upper, e lower side of 

 leaf, nat. size, f lower side of leaflet, X 2; g mature pods,| nat. size, h dehiscence of Sophora 



pod, schematic cross section. 



We cannot, I think, simply use the name 5. tetraptera for the Chilean 

 plant. Philippi called it Edw. macnabiana Grah. Edinb. New Phil. Journ. 26 

 (1839) 196 (figured in Bot. Magaz. 66, pi. 3735). Graham's description was 

 based on a tree cultivated in the botanic garden of Edinburgh; it is not stated 

 where the seeds came from. The main difference between this and the New 

 Zealand plant was that the former »flowered when in full leaf* and was more 

 hardy. I have seen Sophora in flower in some places in south Chile. The 

 leaves are generally, but not always, shed before the flowers open, and the 

 new foliage gets developed with the fruit, — I cannot find that Edw. cassioides 

 Phil, is different. 



The pod of Sophora has an interesting structure. CHEESEMAN, Manual 122, 



