APPENDIX 471 



the seed dried after swelling for germination with those of the same 

 individual seed in the resting state, but only with the average for a 

 number of resting seeds. We are indeed merely comparing an average 

 with an average. Then we have seen in Note 2, A that impermeable 

 seeds when dried after swelling for germination are heavier, and perme- 

 able seeds lighter than in the resting condition. Still, the results given 

 below will tend to emphasise the view that the swelling of germination 

 is essentially a process of water-absorption. This matter is dealt with 

 again in discussing the general mechanism of the shrinking and swelling: 

 processes in Chapter IX. 



NOTE 4 (p. 43). 



Experiments of Victor Jodin on the germinative minimum of Peas (Pimm 

 sativum) y from " Les Annales Agronomiques" October 1897. 



THE peas were placed on a support of platinum suspended in air 

 saturated with water-vapour. By these means the water-vapour was 

 condensed in fine drops, and thus communicated to the peas. A similar 

 experiment under the same conditions, but in which the peas were 

 placed on a support of cork, gave no germination result, even after 

 55 days. In this last case the seeds had absorbed water in varying 

 degrees from 22 to 42 per cent, of their weight, but they had 

 failed to reach the germinative minimum of 67 per cent. This 

 investigator states his results in terms of the hydratation of a seed 

 that is to say, as a ratio of the dry and not of the wet weight ; and it- 

 has been necessary to convert them into percentages of the weight 

 of the resting seed. 



NOTE 5 (p. 69). 



Canavalia ensiformis^ DC. 

 Canavalia gladiata^ DC. 



THE seeds chosen by me as the type of permeable seeds in 

 Chapter IV belong to C. ensiformis, DC., with white seeds. In his 

 Flora of the British West Indian Islands^ Grisebach gives C. gladiata^ 

 DC., as having two forms or varieties : (a) with rufous brown, 

 and (b] with white seeds (C. ensiformis proper). I experimented on 

 both forms, the (a] form possessing seeds with varying degrees of 

 impermeability, the () form with seeds typically permeable. In this 

 work I have applied the specific name of C. gladiata only to the rufous- 

 brown or red-seeded variety, reserving that of C. ensiformis for the plant 

 with permeable white seeds. Professor Ewart in his paper (Proc. Roy. 



