Apeil 7, 1916] 



SCIENCE 



509 



stances, the liquefying agent would be free to 

 act. So far as we know no experiments of this 

 sort have ever been performed, though it may 

 be significant that Setchell failed to find any 

 algse growing at 43°-45° C. 



In a recent paper Osterhout^" advances the 

 hypothesis that substances which increase 

 permeability antagonize those which decrease 

 permeability. He says (p. 256) : 



It seems to the writer that the hypothesis oflEers 

 a rational explanation of antagonism by showing 

 that salts antagonize each other because they pro- 

 duce opposite effects upon the protoplasm. 



The nature of these " opposite effects upon 

 the protoplasm " is an increase or decrease of 

 permeability. Osterhout makes no statement 

 as to the meaning of the term " permeability "' 

 which, without further qualification, is non- 

 committal, nor to the cause of the permeability 

 changes. With these two fundamental gaps 

 in the theory it seems a far cry to a " rational 

 explanation of antagonism." We have empha- 

 sized above that a study of Osterhout's data 

 indicates a direct correlation between de- 

 creased permeability and increased surface 

 viscosity. It seems highly probable, however, 

 that all substances producing an initial de- 

 crease in permeability will, if allowed to act 

 long enough or in sufficient concentration 

 eventually cause an increase in permeability.^^ 

 This conclusion, which we are forced to make 

 from a study of the phenomena of viscosity 

 changes in colloids, complies very well with 

 the experimental data upon permeability 

 changes in both plant and animal cells. 



All physical and chemical agents acting 

 upon a colloidal system influence the state of 

 aggregation of the disperse phase, tending 

 either to increase or to decrease the degree of 

 dispersion. Since we have a colloidal system 

 at the surface of every cell, all physical and 

 chemical agents influence the state of aggrega- 

 tion or its equivalent, the solubility of the 

 surface disperse phases in one of two ways, 

 viz., (1) there may be an increase in the de- 

 gree of dispersion and a corresponding in- 

 crease in the solubility of the disperse phases 



20 Osterhout, loo. cit. 



21 Osterhout calls attention to this fact, but offers 

 no explanation for it. 



and the fluidity of the cell surface, or (2) 

 there may be a decrease in the degree of dis- 

 persion or a decreased solubility of the dis- 

 perse phases which eventually results in a 

 precipitation or coagulation. An " antagon- 

 ism " is to be considered a physiological com- 

 pensation of a force favoring dispersion 

 (solubility) by a second force favoring aggre- 

 gation (insolubility). This relation is recip- 

 rocal. * 



E. A. Spaeth 

 OsBORN Zoological Labokatory, 

 Yale University 



SPECIAL ARTICLES 



NATURAL CROSS-POLLINATION IN THE 

 TOMATO 



Evidence concerning the amount of natural 

 cross-pollination in the tomato has been 

 secured by interplanting two commercial vari- 

 eties of tomatoes, one a standard and the other 

 a dwarf variety. The difference in habit of 

 growth between these varieties is quite dis- 

 tinct in the early seedling stage. The standard 

 is almost completely dominant over the dwarf 

 type of growth. Any pollen from a standard 

 plant fertilizing a dwarf plant should result 

 in a standard plant in the first generation. To 

 test this point a number of dwarf and standard 

 plants were set three feet apart alternately. 

 They were at least five hundred yards removed 

 from any other dwarf tomatoes. These plants 

 were allowed to set fruit normally and seed 

 was saved from the dwarf plants as the fruit 

 ripened. The dates on which the ripe fruit 

 was gathered correspond approximately to the 

 order in which the flowers were fertilized. 

 Seed from these " open-pollinated " dwarf 

 plants was planted in flats in the greenhouse. 

 The number of standard plants which could be 

 plainly distinguished after six weeks' growth 

 was determined and tabulated. 



The approximately two per cent, of crossed 

 plants does not represent all the crossing which 

 might have taken place. Aside from a slightly 

 greater distance, there was an equal chance for 

 the dwarf plants to be fertilized by pollen from 

 other dwarf plants. This crossing would pro- 

 duce only dwarf plants, and hence would not 

 show. 



