94 BOTANY OF SOCOTRA. 



467) in a variety cultivated in Paris as Punica granatum flavum, in which 

 three tiers of carpels are found, and these are primarily concentric. 



Now in our Socotran plant there is but one row of five to seven carpels, and in 

 each carpel there are numerous ovules which are arranged over the floor of the 

 loculus. In fruit it is found that they are spread over the walls of the loculi to 

 some extent, the base of the ovary, as it were, having grown upwards, just as 

 it does in many species of Mesembryanthemum (Eichler Bliithendiagr. ii. 123, 

 f. 46). There is no trace of a second whorl of carpels ; the only row present 

 is that which becomes uppermost in the flowers with two and with three tiers; 

 and it thus appears that we have in the Socotran plant a simpler condition of 

 fruit of the type Punica than in the well-known pomegranate. 



A frequent supposition regarding the pomegranate is that its fruit 

 structure is a monstrous condition developed in cultivation, and we know it 

 has been in cultivation for a very long period. Wight and Arnott (Prod. 327) 

 indeed suggest " perhaps in a truly wild state the upper or adventitious verticel 

 of carpels may occasionally disappear." But even from the districts where it 

 is presumed to be wild — Persia, Kurdistan, Afghanistan, and Beloochistan — (see 

 Alph. De Candolle Origiue des Plantes Cultivees 1883, 189, for an interesting 

 account of the source and early records of the pomegranate) — the pomegranate 

 has been hitherto reported with the doubly verticillate carpels. Here, however, 

 from Socotra we now have a type with a single carpellary whorl. We know 

 so little of the flora of the adjacent Asiatic continent, and there are so many 

 Socotran plants amongst those known therefrom, that it is not unlikely this plant 

 may be found in that region. But, in any case, this plant having the facies so 

 markedly of the pomegranate, differing indeed, except in fruit only in a few 

 minor technical details, might, I think, be considered the type of the primitive 

 stock whence Punica granatum, as it is known in cultivation, has sprung. A 

 further point of considerable interest in the morphology of the carpels is the 

 almost free condition of the ovary in the flower. In the pomegranate the 

 ovary is inferior. This character, too. leads us back to an earlier stage in the 

 evolution of the type. 



Punica was first placed in the natural system by Jussieu in Myrtacese, and 

 his lead has been followed by many botanists, including Lindley (loc. tit), 

 Meissner (Gen. 107), Endlicher (Gen. n. 6340), Berg (in Mart. Flor. Bras. xiv. 

 1. 514), Baillon (Hist, des PI. vi. 330, 378), and Eichler (Bliithendiagr. 

 ii. 488, some considering it deserving of a tribal distinction, others not con- 

 ceding this. Again, Don (in Edin. New. Phil. Journ. 1826, 134) questioned 

 Jussieu's allocation, remarking that the character of the fruit has been quite 

 misunderstood (though his explanation which follows is indeed the most absurd 

 of any that have been advanced), and proposed to make of it a new family, 

 Granatese, of which the affinities he states are uncertain. This family of Don 

 has been kept up by De Candolle (Prod. hi. 3), Martius (Mat. Med. Bras. 50) 



