March 21, 1913] 



SCIENCE 



431 



ing solution can adhere. Tissues with a 

 spongy structure can not be used for exact 

 measurements. Siebeck' has recently been 

 able to carry out a series of experiments on 

 the kidneys of the frog. The measure- 

 ments were not as satisfactory as in the 

 ease of the red blood corpuscles and the 

 muscle, but sufficiently accurate to leave 

 no doubt of the main result, which is as 

 follows : In solutions of NaCl, NaBr, LiCl, 

 and LiBr, KCl, CaCl^, MgCL and cane 

 sugar equi-molecular with a .7 per cent. 

 Einger solution the kidneys neither take 

 up nor lose water. Siebeck points out 

 that these solutions have no other quality 

 in common except the same number of 

 molecules, and that hence the osmotic 

 pressure determines the exchange of water 

 between the kidneys and the surrounding 

 solution. Mathews's experiments indicate 

 that the same is true for the living nerve. ^ 

 According to Meigs^ smooth muscle 

 preparations do not behave as if they were 

 surrounded by semi-permeable membranes ; 

 but the smooth muscle of the stomach 

 which he used in his experiments can not 

 be obtained in as natural a condition as 

 that in which striped muscle or blood cor- 

 puscles or the kidneys are available. It 

 would be of interest to repeat these ex- 

 periments on smooth muscle, which can be 

 obtained with as little alteration of its nat- 

 ural surface, as, e. g., striped muscle or 

 kidneys or red blood corpuscles. 



in. ABSORPTION OP WATER BY COLLOIDAL 

 SOLUTIONS 



The natural media surrounding animal 

 tissues are solutions which contain dis- 



' Siebeck, Fflii-ger's AroMv, Vol. 148, p. 443, 

 1912. 



*A. P. Mathews, Am. Jour. Physiol., Vol. II., p. 

 455, 1904. 



'Journal of Exper. Zoology, Vol. 13, p. 498, 

 1912. 



solved proteins in addition to certain salts. 

 Some authors seem to take it for granted 

 that the force by which such protein solu- 

 tions or colloidal solutions in general ab- 

 sorb water is not determined by Avo- 

 gadro 's law, and some go so far as to state 

 that the regulation of the distribution of 

 water between cells and blood is solely 

 determined by the proteins and not by the 

 salts. Such ideas are contradicted by the- 

 ory as well as by experiment. We can 

 show by the use of living cells as osmom- 

 eters that colloidal solutions behave ex- 

 actly as Avogadro's and van't Hoff's laws 

 demand. 



Avogadro-van 't Hoff's law demands that 

 a grammolecular solution of any dissolved 

 substance contain the sam.e number of 

 molecules — leaving aside temporarily the 

 facts of dissociation — and this number N 

 is Avogadro 's constant. As is well known, 

 the value of N has been determined for 

 molecules with an astonishing degree of 

 agreement by the most diverse methods. 

 Thus Eutherford found 7^ = 62.10- by 

 counting the a particles given off by a 

 given quantity of radium per second; and 

 Dewar as well as Boltwood obtained iden- 

 tical values, 71.10", by a similar method. 



A method of calculating N from the 

 constants of heat radiation gave, according 

 to a theory of Lorentz, A^ = 62.10". Lord 

 Eayleigh determined the number N of 

 molecules in a grammolecule from the dif- 

 fusion of light from the sky and found N 

 about 70.10".i'» 



It would seem a priori perhaps doubtful 

 whether or not the suspended particles of 

 a colloidal solution behave like molecules. 

 Perrin undertook to solve this problem. 

 He points out that van't Hoff's law (for 

 properly dilute solutions) is applicable, no^ 



" These data are taken from Arrhenius, ' ' The- 

 ories of Solutions," New Haven, 1912, pp. 25 and 

 26. 



