OF THE INFLORESCENCE. I'yg 



t. 22(3, Adoxa MoschatelUna, t. 453, and Gom- 

 phrena globosa, the Globe Amaranthus of the gar- 

 dens. 



Perhaps the iniiorescence of Sanguisoi^ha offici- 

 7ialis, t. 1312, might be esteemed a capltulum^ be- 

 cause its upper flowers come iirst to perfection, as 

 in Adoxa, which seems contrary to the nature of 

 a spike ; but it does not appear that all capitate 

 flowers expand in the same way, and Sanguisorba 

 canadensis has a real spike, flowering in the usual 

 manner^ from the bottom upwards. So Allium de- 

 scendensy Curt, Mag. t. 2ol, opens its upper, or 

 central, flowers first, contrary to the usual order in 

 its genus ; both which instances prove such a diver- 

 sity to be of small moment. 



Umbella, an Umbel, for which some authors retain 

 the .obsolete old- English name of Rundle. In this 

 several flower-stalks^, or rays, nearly equal in length, 

 spread from one common centre, their summits 

 forming a level, convex, or even globose surface, 

 more rarely a concave one. When each ray is 

 simple and single-flowered, it is called a simple um- 

 bel, y.' 136, as those of Allium uj^simim, Engl. Bot. 

 t. 122, Ivy, t. 1267, Primula Te?is, t.'S, Jai^inosa, 

 t.6, elatiory t. 513, and Eucalyptus rewi'iferay 

 Exot, Bot. /. 84. In a compound umbel each ray 

 or stalk mostly bears an umhcllula^ or partial umbel, 

 as Athamanta Libanotis^ Engl. Bot. t. 138. This 

 N 2 



