6 TRANSACTIONS LIVERPOOL BIOLOGICAL SOCIETY. 
sense (i.e., a floating magnet brought into the East-West 
position took a very much longer time to return to the 
North-South position than in pure water), although the 
viscosity of the solutions remote from the surface differed 
very little from that of water.  Bile-salts remained 
permanently mobile. Sodium oleate gave varying results, 
possiblv due to impurity or chemical change, and requires 
further investigation. 
When the surfaces are rigid it is clear that solid 
particles previously in solution are present there in 
sufficrently close contact to exhibit the properties peculiar 
to matter in the solid state. When the surface exhibits 
‘** Plateau viscosity,’’ this condition would be adequately 
explained by the assumption that particles are present on the 
surface which lower its tension in some rough proportion to 
their number on a given area, and that the rotating magnet 
sweeps them up in greater density immediately in front of 
it and leaves the area behind it for a short time 
comparatively empty. If no true increase of surface viscosity 
accompanies the Plateau viscosity {and experimental 
difficulties make it almost impossible to decide this point), 
it must be concluded that the particles are either liquid, or 
are solids everywhere surrounded and isolated by intervening 
liquid. The complete absence of rigidity or ‘‘ Plateau 
viscosity ’’ on solution of bile-salts I must leave unexplained. 
The possibility of obtaining mechanical surface aggregates 
by so simple a sweeping-up process as that employed in the 
burette method is only intelligible on the assumption that 
the air surface of solutions which yield solid mechanical 
aggregates, no matter whether showing rigidity or Plateau 
viscosity or an apparently normal limpidity, becomes more | 
or less thickly studded with relatively large solid particles 
which have passed spontaneously out of solution in so far 
that they have come into contact with both air and water. 
