456 KEY. G. HENSLOW ON A PROLIFEBOTTS 



Amongst the flowers borne by the uppermost part of the inflo- 

 rescence there were the following variations : — 



Flowers with the corolla yellow, stamens abortive, with small 

 oblong anthers and centre axis prolonged (PL XVI. fig. 2). 



Flowers with the corolla virescent, stamens abortive, with reni- 

 form apparently one-celled anthers, and axis prolonged and bear- 

 ing a tuft of linear leaves (PL XVII. figs. 1 and 4). 



Both of the above kinds occurred on the higher regions of the 

 upper part of the stem. 



On a lower part of the terminal portion were flowers with very 

 small hairy petals, the pistils terminating above with two linear 

 processes, and with the ovules visible within the ovary (PL XVII. 

 fig. 2). 



At a still lower position were flowers with abortive stamens, 

 while the central axis bore buds below the malformed pistil, 

 now elevated above them (PL XVII. fig. 3). 



The sterility of the anthers is a phenomenon which likewise, 

 but not always, occurred in species of Anagallis arvensis, L., de- 

 scribed by Dr. Marchand* the flowers of which were virescent 

 and proliferous ; but the stamens were never free from the petals 

 or foliar organs which represented them. 



The ovary also of this plant exhibited somewhat similar differ- 

 ences in degrees of prolification, as may be seen by studying the 

 Plate together with pp. 164, 165 (Z. c). 



The chief peculiarity of the flowers on the lower part of the 

 inflorescence consisted in the much enlarged ovaries, which were 

 composed of two carpellary leaves adherent down the median lines 

 to a vegetative axis occupying the place of the axile placenta. 

 In many the ovaries were completely closed, with a short pointed 

 style above (PL XVII. fig. 7). Others had opened near the 



n>lT Yvn ° ne 8lde ' WhenCG a tuft ° f minute leaves P rotruded 

 (1 1. XVII. fig. 8). In others the carpellary leaves had opened 



on both sides, or rather there was an absence of cohesion over the 



summit, so that the leafy axis thrust out its two pendulous leafy 



tufts, one on each side (PL XVII. fig. 9). In other instances the 



two carpels assumed the form of narrow leaves with widening 



• 'Adansonia/ iy. p. 150. See also a paper on a somewhat similar monstro- 

 sity of Lmana Elatine, Desf., by the same author, Bull, de la Soc. Bot. de la Fr. 

 18,9, p. 107 The reader may also consult Dr. 0. Cramer, ' Bildungsab- 

 weichungen bei ein.gen wichtigeren Pfianzenfamilien,' &c, Heft 1, tabb. ix., 

 x., xii. 



