1909.] Coverings of the Seeds o/Hordeum vulgare. 



93 



Summary of Conclusions. — The investigation of the selective properties of 

 the semi-permeable seed-coverings of H. vulgare described in this paper should 

 be regarded as pioneer work only ; much further study is required in order 

 to explain the varying actions of the seeds in the presence of different solutes 

 in aqueous solution. 



At present, the general trend of the evidence tends to show that solutions 

 of the solutes which diffuse readily through the seed-coverings differ in some 

 essential manner from solutions of non-diffusible solutes, although the nature 

 of the difference remains unexplained. The results of some of the earlier 

 experiments described above appear to support the view that the property of 

 diffusion is intimately associated with a low degree of " ionisation " of the 

 solute ; yet the conspicuous instance which has been noticed of the ready 

 diffusibility of trichloracetic acid, a highly " ionised " acid, tends to show 

 that such correlation, if it exist at all, is not an intimate one. Further, the 

 view does not appear to be favoured by those experiments which have 

 demonstrated that certain non-electrolytes, such as ethyl alcohol, are readily 

 diffusible, whilst others, such as glycerol, are non-diffusible. 



In connection with the same question, it seemed possible that differences 

 in the surface tension of solutions of diffusible and non-diffusible solutes 

 might perhaps be associated in some way with the different behaviour of the 

 two classes of solutions towards the seed-coverings ; but it appears from 

 a study of the surface tensions of the two classes of solutions that there is 

 no such intimate connection between them. Neither can any indication be 

 found that viscosity is associated with the manner in which diffusible and 

 non-diffusible solutes behave differently towards the seed-coverings. 



The only explanation of the observed difference in activity of the two 

 classes of solutions which at present suggests itself as a working hypothesis 

 is, that some unrecognised peculiarity in the manner in which the molecules 

 of the two classes of solutes are combined with the molecules of the solvent 

 water may constitute the factor which orders their different behaviour with 

 respect to the seed-coverings. This hypothesis appears to be supported by 

 the experiments which demonstrate that, whereas readily diffusible solutes 

 enter the seed together with a large amount of water, seeds placed in 

 solutions of non-diffusible solutes absorb water with some difficulty. 

 Moreover, the observation that an aqueous solution of alcohol diffuses 

 readily through "the seed-coverings which are impervious to this solute 

 in the anhydrous state, appears to show that some form of combination of 

 solute and water is necessary to condition diffusion of the solute through the 

 seed-coverings. 



