286 BOTANY OF THE LIVING PLANT 



the carpel of Anemone only a single ovule matures, though several are initiated 

 but remain vestigial (Fig. 22S, a). In Rue and Buttercup only a single ovule 

 is present (Fig. 228 b, 229). On the other hand, the number of carpels in the 

 Helleboreae is about five ; but a much larger number are present in Anemone 

 or Buttercup ; the more numerous carpels make up for the reduction of the 

 ovules to one in each. Similarly in the Leguminosae a primitive type is seen, 

 with many ovules in each carpel in the Pea or Bean. But Copaifera contains 

 but two ovules, of which one only ripens into an unusually large seed (Fig. 225, 

 p. 282). Again, in the Rosaceae, in Spiraea or Cydonia each carpel is many- 

 ovuled ; but in the Potentilleae, Rubeae and Roseae each has only one. 

 The loss of number is, however, made up by the greater number of carpels. 

 In the Pruneae, on the other hand, the single carpel contains two ovules, of 

 which usually only one matures ; but it grows into a large kernel. The 



Fig. 229. 



Fig. 227. Follicles of .\coiute. (.\fter Figuier.) 



Fig. 228. a, Carpel of .Anemone, with aborti\'e o\'u1gs. b. Carpel of Thaliciru}}!. 



(.After Prantl.) 

 Fig. 22<i. .\ehcne or nut of Buttercup. (.After Figuier.) 



gynoecium of the Oak is trilocular, with two ovules in each ; but only one of 

 the six ovules matures into the seed in the large acorn. The Coco-Nut again 

 shows thi"ee loculi, but onl)' one matures its single very large seed. A similar 

 condition on a smaller scale is seen in Valerianaceae, while the gynoecium, with 

 a single loculus and a single ovule in each flower, becomes the rule in the Teasels 

 and Composites: these may be held to be examples of an extreme state. 



Such evidence clearl)' indicates a progressive reduction in number of ovules 

 produced and matured in the cases quoted. It often goes with increase in 

 size of the individual seed, as in the Plum, Oak, and Coco-Nut. which gives 

 greater certainty of successful germination. But some families show the 

 converse, viz. that there has been a progressive increase in number of the 

 .seeds, though tlie individual size of flic seeds is diminished. The most marked 

 examples are seen where some irregular form of nutrition makes germina- 

 tion hazardous. For instance, in the mycorhizic Orchidaceac, where over 

 a million seeds have been estimated as the produce of a single capsule. In 

 many of the mycorhizic Ericaceae the seeds are numerous ; and the same is 



