COLLOIDAL BEHAVIOR OF PLANT PROTOPLASM. 169 



newed by metabolism instead of being furnished as in the experi- 

 ments described. 



Of the various suggestions which might be offered in explana- 

 tion of the reaction, the most plausible one seems to be one in which 

 it is assumed that the replacement of the glycocoU solution would 

 result in the formation of a theoretically possible glycocoll agarate, 

 the bulk of which might be greater than the total of its separate 

 components, and hence the combination would result in an immediate 

 further swelling. The slow diffusion of this substance out of the 

 sections would lessen or prevent the absorption or penetration of 

 the glycocoll solution from the outside so a shrinkage would result. 

 In the case of the hydroxides and mixtures containing gelatine the 

 combination most probably would be that in which a potassium gela- 

 tinate would be formed and its slow diffusion would be accompanied 

 by a shrinkage. In any case the changes in volume seem to have 

 escaped observation hitherto and to be of such range as to have 

 significance for the mechanism of the cell. 



Structure and Heterotropic Swelling gf'Collgidal Mixtures. 



That dried sections of colloids do not show equivalent expansion 

 in all directions due to the development of structure in disiccation 

 has long been known and has been variously discussed in previous 

 articles. Extreme differentiation is shown by agar, plates of which 

 may be so dried that they increase but 2-4 per cent, in length and 

 width while swelling 3,000-4,000 per cent, in thickness. Gelatine on 

 the other hand may be cast in such form that it increases 10 to 60 

 per cent, in length and width while hydrating 1,000 to 3,000 per 

 cent, in thickness. Mixtures of agar and gelatine do not show 

 more than 10-16 per cent, increase in superficial measurements. 

 No accurate measurements have been made, but the data in Tables 

 I. and II. suggest that the swelling in the axes in which the colloid 

 did not shrink when desiccating may be modied in a distinct manner 

 characteristic of the substances acting upon the colloid. Thus the 

 relative increase in volume is greater in a mixture of i part agar 

 and 3 of gelatine in hydroxide than it is in acid, and other differ- 

 entiations may be found by inspection of these tables. 



