October 20, 1916] 



SCIENCE 



565 



completion and will be ready for occupancy 

 about December 1, when Dr. Arthur A. ISToyes 

 will go to Pasadena, where from now on he will 

 spend half of each year. James H. Ellis, 

 Ph.D. (Mass. Inst.) has been appointed re- 

 search associate in physical chemistry in the 

 college. 



The appointments of Drs. Edward H. 

 Nichols and Charles A. Porter as clinical pro- 

 fessors in the Harvard Medical School have 

 been confirmed by the university's board of 

 overseers. Both men formerly held positions 

 as associate professors. 



Harold Veatch Bozell has been appointed 

 assistant professor of electrical engineering in 

 the Sheffield Scientific School of Tale Univer- 

 sity for the college year 1916-17, for which pe- 

 riod he has secured leave of absence from the 

 University of Oklahoma, where he is dean of 

 the school of electrical engineering and pro- 

 fessor of electrical engineering. 



Professor 0. 1ST. Haskins, of Dartmouth 

 College, has been promoted to a full professor- 

 ship of mathematics. Drs. R. D. Beetle and 

 P. M. Morgan have been promoted to assistant 

 professorships of mathematics. 



Roy G. Hoskins, associate professor of 

 physiology in the Northwestern University 

 Medical School, has been promoted to be full 

 professor and head of the department. Virgil 

 Ernest Dudman, M.D., interne in Cook County 

 Hospital, Chicago, has been elected assistant 

 professor of hygiene and director of student 

 health. 



H. H. Bunzell, Bureau of Plant Industry, 

 has been appointed assistant professor of bio- 

 chemistry at the University of Cincinnati Med- 

 ical School. 



Dr. A. A. Bennett, of Princeton University, 

 has been appointed adjunct professor of mathe- 

 matics at the University of Texas. 



O. P. Burger has been appointed instructor 

 in plant pathology in the Graduate School of 

 Tropical Agriculture of the University of Cali- 

 fornia at Riverside, and Alfred Free Swain, 

 formerly of Montana State College and of 

 Stanford University, assistant in entomology 

 there. 



DISCUSSION AND CORRESPONDENCE 



COLLOIDS AND NEGATIVE SURFACE TENSION 



In a review of Professor Pischer's transla- 

 tion of Wo. Ostwald's " Handbook of Colloidal 

 Chemistry " which recently appeared in this 

 journal, 1 Professor W. A. Patrick states that 

 the existence of negative surface tension which 

 is assumed by Ostwald is contrary both to ex- 

 perimental evidence and to the fundamental 

 ideas of surface tension. Although the pres- 

 ent writer does not agree with all of Ostwald's 

 energetic considerations, he wishes to point 

 out that the existence of negative surface ten- 

 sion under certain circumstances is not only 

 supported by a vast body of experimental evi- 

 dence but is necessitated by the thermodynamic 

 theory of the stability of colloidal solutions. 



Surface tension may be defined in the usual 

 way as the work which has to be done in order 

 to increase the surface in question by one 

 square centimeter, this increase in surface 

 being carried out of course reversibly and iso- 

 thermally. This work, however, and hence 

 also the surface tension, may be either posi- 

 tive or negative. 



If for a given two-phase system the surface 

 tension at the boundary between phases is 

 positive, then a positive quantity of work will 

 have to be done in order to increase this sur- 

 face, and such an increase in surface will be 

 accompanied by an increase in the free energy 

 of the system. Since all spontaneous changes 

 in a system must be in the direction of de- 

 crease of free energy, these systems with posi- 

 tive surface tension if left to themselves will 

 automatically decrease the surface between 

 phases. Thus, for example, in the case of 

 finely divided crystals of copper sulfate in con- 

 tact with a saturated solution, we have a sys- 

 tem in which there is positive surface tension 

 at the boundary between the phases, and if this 

 system is left to itself there will be a spon- 

 taneous decrease in surface, the smaller crys- 

 tals going into solution and precipitating on 

 the larger until finally we have all the solid 

 copper sulfate in one large crystal, this being 

 the condition of smallest possible surface. 



If, on the other hand, we have a two-phase 



i Science, 43, 747, 1916. 



