530 



REPORT — 1901. 



Before commenciug operations with the colloidal solution, experiments were 

 made with a solution of barium chloride, so as to obtain a check on the results ; 

 the agreement is satisfactory. 



I. 



II. S7>^, 



T 

 19-4° 

 19-4° 

 20-67° 

 2Cr67° 



Hydrostatic 

 method 



1-02677 



1-02928 



T Pyknometer 



20° 1-02681 1 



20° 1-02685 f ™^^°' 1'02683 



20^4^^ 1.029.3] 



20-47° 1-02930 



The difference in the first case is 6 in 102,000; and in the second, 2 or 3 in 



,000. 



With colloidal solution of arsenious sulphide, the data were : — 



100,000 



1. 



II. 



III. 



IV. 



T 



21-2° 

 21-2° 

 20-7° 



20-7° 



20-75° 



20-75° 



20-80° 



20-80° 



Hydrostatic 

 method 



1-01187 

 102323 

 1-02272 

 1-01129 



T 

 21-2° 



20-80° 



Pykno- 

 meter 



101192 

 1-01193 

 1-02330 

 1-02329 

 1-02276 

 1-02277 

 1-01135 

 1-01134 



0-00005 



0-00006 

 000005 

 000005 



The solution IV. was prepared by diluting III. ; the others were all specially 

 prepared ; they contained arsenious sulphide of such a degree of subdivision that 

 the particles polarised liglit, but were invisible under a magnification of 1,000 

 diameters. 



The influence of error in weighing is such that an error of 1 mgr. in thev/eight 

 of the body in air or in solution would bave caused an error of 2 units in the fifth 

 decimal place : and 1° in reading temperature would have made an error of the 

 same magnitude. 



It will be noticed tbat the apparent density with the pyknometer always 

 exceeds that with the float by 3 to 6 units in the fifth place of decimals ; i.e., by 

 3 to 6 parts in 100,000. It is probable that this is due to some unapplied correc- 

 tion ; but it is not easy to allow for it. It may, we think, be taken as proved 

 that colloidal arsenious sulphide of the state of subdivision used, exerts hydro- 

 static pressure as if it were a liquid ; at any rate, it behaves as if it were in true 

 solution like barium chloride. 



It has long been the custom to determine the density of milk, which contains 

 suspended fat globules, by means of a lactometer, which involves a hydrostatic 

 method. The experiments cited show that this custom is justifiable. 



6. The Freef-i7i(/ Points of certain Dilute Solutions. 

 By E. H. Griffithb, F.R.S. 



6. The Buildings of the National Physical Laboratory. 

 By Dr. R. T. Glazebrook, F.R.S. 



FRIDAY, SEPTEMBER 13. 

 The Section was divided into two Departments. 



