June 15, 1^93] 



NATURE 



153 



means of a wire previously heated to redness, the water- 

 surface still remaining normal. If tallow was to be tried, 

 I left several Iragments floating on the water for a short 

 time. When the anomalous surface had reached a suffi- 

 cient length, the floating wire was put upon it about half- 

 way between the partition and the end of the trough. 

 Then the sliding weight of the balance was displaced 

 successively along intervals of the scale, corresponding 

 to equal differences of tension, and after each displace- 

 ment one determined the length of surface, at which, 

 under continued contraction, the disk or ring broke off. 

 From these lengths the relative contaminations were 

 afterwards calculated. 



Thus I obtained the following results, T denoting rela- 

 tive surface tension and R relative contamination : — 



It did not influence the results if I used poppy-oil 

 instead of olive-oil, or tallow of various provenience. 



The decrease of tension was rapid, and nearly propor- 

 tional to the increase of relative contamination, till the 

 value 0-82 in the case of oiV or 079 in the case of tallow, 

 was attained. At this point a sudden change occurred, 

 and the further sinking took place very slowly. At the 

 same time the " solution current " of floating tallow frag- 

 ments showed a sudden lessening. 



Under continued contraction the water surface at last 

 appeared turbid, and the lowest tension I could attain in 

 this way was about 063 with oil ando68 with tallow. 



The method described is still somewhat imperfect, in- 

 asmuch as the water particles in close pro.ximity to the 

 sides of the trough did not participate in the movement 

 of the rest of the surface, indicated by the displacement 

 of the wire-mark. Therefore the results were checked 

 by another method. 



The whole surface of the trough was rendered anoma- 

 lous by means of weak solutions of oil or tallow in 

 benzol, for which purpose 23 drops of the oil solution, or 

 13 of the stronger tallow solution were required. If the 

 contamination be a little too great, the normal tension 

 may be easily restored by immersing small strips of paper. 

 Then part of the surface was cleansed by shifting the par- 

 tition from the end towards the middle of the trough 

 about 10 cm., and one drop of the solution being 

 evaporated on the newly-formed surface, the partition 

 was removed and the ten -ion measured. The increase 

 of relative contamination thus added by e.ich drop was 

 respectively }.-^ and jJ.j. The means of all observations 

 made in this manner were as follows : — 



Provenc; Oil. Tallow, 



The results agree tolerably well with those obtained by 

 the first method, and show still more clearly, that the 

 tension in the beginning of the anomalous state may be 

 appro-ximately expressed by 



T„ - T = k (R - I), 



T„ denoting the normal surface-tension and k a con- 

 stant which is, in the case of oil, 0-60, and in the case of 

 tallow, 0-54. 



The course of the tension of a surface contaminated 

 by oil may be more clearly seen from the following curve : 



When the water-surface is not contaminated by pure 

 grease, but by any other substance, as soap, resin, pal- 

 mitic acid, the tension behaves quite differently. 



If the justly anomalous surface be contracted, the 

 tension at first sinks rapidly, but gradually begins to rise 

 again, while the area of the surface remains constant ; 

 the latter being re-extended, the normal condition is at- 

 tained at a shorter surface than before contraction. 

 When the surface then is left for some time in the normal 

 state, the length of the anomalous surface increases 

 again. 



Therefore the dilTerence of tension produced by varying 

 the area in a given ratio depends very much upon the 

 time required for the contraction or extension. 



The deeper the tension is lowered the stronger be- 

 comes its tendency to rise, till at last a further sinking 



' When ordinary o iveoil w.'is used the value in question w-is 0-78. 



NO I 2 ;;^ vni,. aSl 



only can be observed during the motion of the partition, 

 and this even is the case with grease at higher degrees of 

 contamination. A sudden change of direction at a cer- 

 tain lowered tension I could perceive in no other curve 

 than that of pure grease. As an example may be ex- 

 hibited the curves of soap and mastic, when the 

 contraction was as quick as possible. Mastic was intro- 

 duced by means of benzol, soap directly by contact. 

 On strongly contracted surfaces, every substance gives 



