ioi4] DARSIE, ELLIOTT & PEIRCE— GERMINATING POWER 103 



from Kew in 1856 for the University Gardens, but these not being 

 ready at the time expected, the seeds had been put away in a dark 

 dry closet and had remained there unopened for upward of 50 years. 

 In addition, he examined some ten-year-old seeds from Sydney and 

 Adelaide, and more from the National Herbarium, making alto- 

 gether nearly 3000 tests. He first tried germinating the seeds in 

 soil, but finding this unsatisfactory, he soaked the seeds and then 

 placed them on moist filter paper in glass dishes and set them in a 

 germinator. Seeds which did not swell after one or two days in 

 water were either filed or treated for 15-90 minutes with concen- 

 trated sulphuric acid, that is, until the cuticle was dissolved away. 

 Adansonia, for example, required almost 6 hours' treatment of this 

 sort. The seeds were then washed and thereupon swelled readily. 

 Ewart adds to his own long list by including some of the results of 



Becquerel, Nobbe, De Candolle, Girardin, Darwin, Duvel. 

 Romanes, Peter, Berkeley, and others, making a list, therefore, 

 which comprises about 4000 species. He too found that the 

 majority of those seeds which retain their germinating power for 

 the longest term are members of the Leguminosae, and are generally 



hard-shelled. 



chemical conditions prevailing in dormant 

 )f opinion corresponding to the paucity of 



knowledge. Ewart, for example, states that molecular 



may 



> 



ements continue until finally the seeds no longer retain 

 resume active life. Whatever these molecular changes 



it is obviously impor- 

 tant, for both theoretical and practical reasons, to determine the 

 conditions in the seeds themselves and the means of maintaining 

 and bettering these if possible. In this paper, however, we are 

 concerned with the results of these conditions rather than with the 

 conditions themselves. 



Until recently, so far as we know, tests of the longevity or via- 

 bility of seeds have depended upon the percentages of actual ger- 

 minations in prepared beds. Such tests are simple enough in the 

 case of seeds which germinate quickly, and in these cannot be 

 improved upon. When, however, two weeks or more must elapse, 

 even under the most favorable conditions, before one may know the 



