22 BOTANICAL GAZETTE [july 



The attempt to measure the internal forces of seeds with semi- 

 permeable coats by means of osmotic solutions was based upon 

 this conception of the moisture relations, and on the assumption 

 that the total osmotic pressure of the solution is transmitted through 

 the membrane as force when pitted against the internal forces of 

 the seed through the agency of the semipermeable coat. The 

 results of this attempt have been highly satisfactory, and I can 

 see no good reason for doubting that the internal forces of air-dry 

 seeds are approximately equal to 900-1000 atmospheres. 



The vapor pressure method,, using sulphuric acid, involves 

 another assumption whose validity may be a little more question- 

 able. It is generally admitted that the osmotic pressure of a solu- 

 tion can be calculated with some degree of accuracy from its vapor 

 pressure. I have assumed in addition that the vapor pressure of 

 the seed hydrogels measured against the vapor pressure of strong 

 solutions can be used as a measure of the internal forces of the seed. 

 The assumption seems reasonable enough, but it would be difficult 

 to offer definite proof of its validity at present. 



The principal difficulty with the vapor pressure method is the 

 uncertainty as to the osmotic pressure of sulphuric acid. It may 

 seem quite unwarranted to some even to think of estimating the 

 osmotic pressure of strong sulphuric acid by means of a formula, 

 knowing as little as we do of all the factors which enter into the 

 problem in this case, and knowing that these formulae have all 

 been developed for the dilute "ideal" solutions. I realize fully 

 the danger of basing any conclusions upon results obtained by such 

 precarious methods, and offer the results merely for their suggestive 

 value. The formula I have used for calculating the osmotic pres- 

 sure from the vapor pressure is that given by Walker (35), with 5 

 representing the density of the solution rather than that of the 

 solvent. This is necessary in order to make the formula fit a 

 concentrated solution with high density. This same formula has 

 been used by others for the same purpose, as by Drabble and 

 Drabble (12). 



Attention is called to a difference in the formulae given by 



Nernst (24) and Walker. In Walker's discussion, —r- repre- 



