MB. G. KTOG ON A SPOET IN PARITIUM TRICUSPE. 103 



There is a very striking resemblance indeed between the foliage 

 of this sport of P. tricuspe and the normal form of the allied 

 species P. tiliaceum, A. Juss. The flowers of the two are also 







much alike except in the subsidiary matter of colour. Those of 

 P. tricuspe are of a bright yellow with a purple eye. On wither- 

 ing, the yellow changes into a dusky purplish brown, which is 

 very nearly the colour of the flower of P. tiliaceum from the 

 first. 



Two points I have omitted to notice ; and these are that the 

 plants of P. tricuspe in the Saharunpore Garden which carry 

 leaves of the normal form incline to the habit of trees, whereas 

 those with the abnormal form of leaf (and in this they resemble 

 the original sport) are more bush-like, and so approach in habit 

 all the plants of P. tiliaceum in the garden. 



Note. — The sport of which an account is given by Dr. King (and 

 specimens illustrating which are placed in the Kew Herbarium) 

 was originally noticed by Mr. Bell, ' Trans. Bot. Soc. Edinburgh,' 

 vol, vii. (1863) p. 565. It is quoted by Mr. Darwin, l Variation of 

 Plants and Animals under Domestication/ vol.i. p. 377, but (by a 

 clerical error) as Cistus instead of Hibiscus tricuspis. It may be 

 noted that H. tricuspis (Paritium tricuspe) is a native of the South- 

 Sea Islands, from which, according to Roxburgh, it was introduced 

 into India. H. tiliaceus, on the other hand, is widely distributed 

 throughout the tropics. The facts described by Dr. King point 

 in the direction of H. tricuspis being a local form derived from 

 II. tiliaceus , to which in the Saharunpore Garden it again re- 

 verted. These are facts w T hich appear to show that plants with 

 undivided leaves tend to develop foliage which is more or less 

 divided when they form part of the vegetation of islands. This 

 seems to be probably the case with the island of Rodriguez, ac- 

 cording to Mr. Balfour.— W. T. T. D. 



» 



* * 



