ON THE LIMITING THICKNESS OF LIQUID FILMS. 
655 
When the tubes were placed in position the colours of the bright transmitted bands 
were, owing to the large number of films, very vivid, and the dark bands were very 
obscure. The fringes which had previously been made vertical were seen crossing the 
bright bands in a more or less sloping direction, but were completely lost in the 
darker portions of the field. When the films had thinned sufficiently to show by 
reflected light the white and black of the first order, the passage from the one tint to 
the other was, as is usual, so sudden as to appear discontinuous. The corresponding 
dark and light transmitted bands were very intense and the boundary between them 
was also perfectly definite and sharp. The fringes visible in the bright portion of the 
field (corresponding to the black seen by reflected light) were for the greatest part of 
their length vertical, but at first they often displayed a very considerable curvature at 
their lower extremities. In such cases they crossed the boundary at a very small 
angle and were lost in the dark band. The direction of the curvature was different 
on different occasions. The accompanying figure is a reproduction of a sketch on an 
enlarged scale made at a time when the phenomenon was very marked. The cause of 
the curvature was evidently an increasing difference of thickness between the films 
in the two tubes in the neighbourhood of the limits of their black portions. 
As all did not thin at precisely the same rate, this limit was in different films at 
different vertical elevations, and it might at first sight seem probable that the curva¬ 
ture was due to the fact that the boundaries between the white and black were in 
the one tube lower than in the other. The interfering rays in a given horizontal 
plane would thus, in the one case, traverse black films only, while in the other they 
would also pass through some white ones, and, as the number of these would increase 
rapidly as the vertical elevation of the plane diminished, a distortion of the fringes 
similar to that observed might have been produced. 
If this had been the true explanation, we should have expected either that the 
intensity of the illumination would have increased gradually on passing from the 
dark to the bright transmitted bands, or that a marked discontinuity would have 
been observed in the curved fringes at the points where the number of the white 
films traversed increased. Neither of these phenomena were observed. The dark 
part of the field was very intense close to the boundary and the curves were 
unbroken. 
Another hypothesis which would serve to explain the phenomenon is that the 
black portions of the films increased in thickness near the junction with the white, 
and that this increase was different in the two tubes. No evidence of any such 
4 p 2 
