THE COMMELTNACE^ OF BEITGAL. 



441 



of its cells contain each two superimposed seeds and dehisce ; 

 the third cell contains one much larger and diflferently shaped 

 seed, and is strictly indehiscent ; it finally withers on the seed so 

 firmly that it can hardly be removed with a knife ; it remains 

 a long time on the peduncle ; and I much suspect some of the 

 species described as 1-seeded. The 1-seeded cell is placed lower 

 down the axis than the other two. 



The two seeds in one of the dehiscent cells fit very closely one 

 upon tte other, and form together a shape mimicking that of 

 the solitary seed. I think it not improbable that some of those 

 who have found but three seeds in C. communis have had not 

 well-ripe specimens, and have counted the two superimposed 

 seeds as one, instead of satisfying themselves as to the number of 



the embryos. 



Pictures of this type of capsule will be found in Webb and 

 Berthelot's 'Canary Isles,' plates 238, 239, and in Wight's 



' Icones,' t. 2065. figs. 7, 8. 



In Sect. I. 5, the capsule is slightly unsymmetrical ; two of 

 the cells containing each one seed dehisce; the third cell is 

 placed slightly lower, is somewhat gibbous, and is strictly inde- 

 hiscent, withering firmly on the seed and remaining a long time 

 on the peduncle. The three seeds are nearly equal and similar, 

 and like the solitary seed in Sect. I. a. The capsule and seeds, 

 at first sight, appear exactly as in Sect. I. a, the two super- 

 imposed seeds sticking a little closer together than usual. 



In Sect. II. the capsule is symmetrical in all respects, contains 

 one seed in each cell ; and all the cells dehisce. The seeds are in 

 shape like the large solitary seed in Sect. I. a. 



Fig. 1. 



Fig. 2. 



Fig. 3. 



Fig. 4. 



Fig. 5. 



