EANJJNCULACEM. 9 



apex of the abundant flesliy albumen. The receptacle swells into a 

 hypogynous disk of little thickness on a level with the insertion of the 

 stamens, and also forms a projecting pad below the base of the calyx. 

 N. GarideUa is found in the countries of the Mediterranean ; it is a 

 herbaceous annual, with erect, angular stem and branches, leaves 

 alternate, pinnately compound, much divided ; flowers solitary, 

 terminal. 



The other Nif/ellce only differ from N. GarideUa in characters of 

 but little importance, such as the larger number of parts in the 

 corolla, androceum, and gynceceum. In fact, if we look at the flower 

 of N. arvensis L. (figs. 20, 21), we see that the perianth is com- 

 posed of five sepals and eight petals — a difference between the number 

 in each whorl which is at first surprising. But on examining the 

 relative positions of the sepals and petals, we see that while the 

 lateral sepals have each a petal opposite them, as in N. GarideUa, the 

 three other sepals have each two petals before them. We must 

 therefore consider that the yigeUce have a corolla of five petals 

 opposite the sepals, of which three undergo reduplication.' The 

 stamens are formed and spirally arranged as in N. GarideUa ; 



Fig. 20. 

 Flower. 



Niyella arvensis. 



Fig. 21. 

 Longitudinal section of flower. 



but the secondary spirals, eight in number, which end opposite the 

 base of the petals, are distinct and very marked; and hence the 



' It even happens that in the Nigella culti- 

 vated in our gardens, all the petals may be 

 doubled, and replaced by as many pairs, each 

 opposite a sepal (see AdanJtonia, iv. 10). But as at 

 the same time the petals have the same singular 

 form as in GarideUa, and are produced, not simul- 

 taneously, but successively in a spiral order like 

 the parts of an androco-um (Payee, Organog., 



248) ; as each of these begins a row of stamens, 

 as in Jiraulhis, &c. ; taking also into considera- 

 tion the opposition of the parts of the so-called 

 corolla to those of the c;ilyx, we are led to think 

 that the nectaries of Sigella, as the older bo- 

 tanists termed them, represent staminodes, not 

 petals — an interpretation that, moreover, in no 

 way aftccts the symmetry of the flower. 



