to germinate, in Seeds belonging to different Families. 45 



we compare the annual, biennial, perennial and ligneous plants. 

 There were in the experiment 357 species, the duration of which 

 is known from botanical books, and 11 of which this is doubt- 

 ful, either in itself or on account of some doubt as to the specific 

 name. The 357 species which may be taken into account class 



themselves thus : — 



Number of Out of 



Total those which 100 species 



Species. number. came up. came up. 



Annuals 180 9 5*0 



Biennials 28 0-0 . 



Perennials 105 4 3-8 



Ligneous 44 3 07 



Total ~357 10 4-4 



Or, 



Monocarpons 208 9 4-3 



Polycarpons 149 7 4-7 



Total 357 10 * 4-4 



These figures seem to prove that the woody species preserve the 

 power to germinate longer than the others, while the biennials 

 would be at the opposite extreme. However, we must observe 

 the small number of species in these two categories, from which 

 it will be concluded that with regard to them the experiment 

 was insufficient. As to the perennial plants compared with the 

 annuals, it seems probable that their faculty of germination is 

 lost rather more quickly. 



Do the large seeds preserve the faculty of germination better 

 than the small ones ? Our experiment can answer this question 

 but imperfectly. In fact, we did not sow very large seeds, like 

 those of the Cocoa for instance ; nor even large seeds, like those 

 of many Palms, certain Leguminosse, Sapotacese or Coniferse. 

 The seeds in the collection were of a mean size, like the Haricots, 

 the seeds of Iris, of Convolvulacese, &c; or small seeds, like 

 those of Composite, Graminacese, Geraniacese, &c; or lastly, 

 very small seeds, such as those of the Poppies, Plantago, Ama- 

 ranthacese, &c. It would be difficult to class all the seeds sown 

 in the different degrees of magnitude; only we may remark 

 among the species which came up a rather large proportion of 

 the mean or small (Doliehos, Malvaceae, Balsam, Acacia, Vicia), 

 while the very small seeds did not come up. We have thus a 

 confirmation of the idea conceived at the time of sowing, namely, 

 that the reproduction of the species with very small seeds is 

 assured by their number rather than by the duration of the ger- 

 minative power. This is not surprising, since the very small 

 seeds have a much larger surface in proportion to their volume, 

 and are consequently more readily penetrated by the variations 

 of temperature or humidity which affect the organs. It is pro- 

 bable also that a much larger proportion of the very small seeds 



