70 BOTANICAL GAZETTE [july 



were evidently absorbed in the required quantities irrespective of what was 

 offered in excess. Those of the second group varied widely, sometimes with 

 an increase of the ion offered in excess, as in the case of Na20, and sometimes 

 with an increase of some other element, as in the case of Si02 in the plots 

 receiving CaCOj and acid phosphate. The soda-potash ratio was subject to 

 extreme variation, but was always greater than i in both tops and roots. 

 There were indications that sodium may partly replace potassium in function 

 in spinach, since the percentages of the two usually varied in the reverse order. 

 The writers suggest that NaCl as a fertilizer for other crops might serve as a 

 potash sparer. There was always more magnesia than lime present, except 

 in the plots receiving a heavy treatment of CaCOj, which suggests the possible 

 practical value of magnesium salts as fertiUzers for spinach. — J. J. Willaman. 



Permeability. — A new working hypothesis as to the nature of permeability 

 and changes in permeability of protoplasm, which seems to the reviewer less 

 objectionable than any yet proposed, is offered by Free.* Protoplasm is 

 considered as a colloidal system of at least two phases, differing from one another 

 mainly in the proportion of water each contains, and arranged as colloidal 

 globules in a colloidal medium. These phases- are supposed to exhibit inter- 

 change of water, so that globules may decrease in size by giving up water to 

 the medium, which gains in size; or, vice versa, globules may increase in size 

 by receiving water from the medium, which thus becomes thinner and thinner 

 as the globules enlarge. 



The medium is considered the important phase from the standpoint of 

 permeability changes, as it is continuous. Anything that can dissolve in the 

 medium should be able to penetrate. Water undoubtedly penetrates both 

 phases. Anything which tends to increase the size of the globules at the 

 expense of the medium is conceived to decrease the permeability of the proto- 

 plasm; conversely, things tending to decrease the globules are conceived to 

 increase its permeability. Semipermeability is related to a very thin medium 

 between the globules. Any reagent increasing thickness of the medium at the 

 expense of the globules should decrease semipermeabiUty if this conception is 

 correct. Antagonism would be explained by the effect of the antagonistic 

 element or ion on the globules, enlarging them so that the medium is too thin 

 to permit entry of the toxic element. As a working hj'pothesis it has some 

 advantages over any other hypothesis which has been proposed. It should 

 stimulate research designed to test its merits, for definite testing seems quite 

 possible. — C. A. Shull. 



Self-sterility. — East and Park have already demonstrated^ that self- 

 sterility in tobacco is heritable, and that cross-sterihty depends upon likeness 



* Free, E. E., A colloidal hypothesis of protoplasmic permeability. Plant 

 World 21:141-150. 1918. 



' BoT. Gaz. 66:461-462. 1918. 



