1871.] Blood-corpuscles through the Walls of the Vessels. 559 



of the laws which are operative in connexion with delicate colloidal films 

 in general. 



The steps, for example, in the production of an ordinary soap-sphere 

 are very remarkable, as exemplifying the power which these films possess, 

 under the influence of progressive cohesion, to perfect any absence of con- 

 tinuity which may exist in their structure. 



The first essential in the process of forming a soap-sphere is the pro- 

 duction upon the mouth of the pipe-bowl of a film stretching evenly across 

 from every point of the circumference. 



The production of this film is a far more complex operation than is 

 generally supposed. 



If for the pipe-bowl we substitute a ring having a diameter of from 12 

 to 18 inches, we are enabled to watch, as the process proceeds, the manner 

 in which the film is formed. 



Having submerged the ring in a solution of soap, we observe, as we 

 gradually raise it out, that its circumference brings up from the liquid a 

 band-like film of a cylindrical or tubular form which is attached to the 

 ring above and the liquid below ; raising up the ring still higher, we find 

 that this annular film contracts in diameter at every part except at its at- 

 tachment to the circumference of the ring, which is of course fixed. This 

 quality of the film to contract between opposing points of extension 

 causes it to take on the shape of an inverted cone with curved sides, the 

 convexities of which are directed inwards. The tendency to assume the 

 inverted-cone shape is further assisted by the fact that the film, in con- 

 tracting, travels inwards upon the surface of the solution towards a central 

 point, so that from the ring downwards to the surface of the solution the 

 diameter of the tubular film is continually decreasing. The shortest 

 diameter is not, however, immediately upon the surface of the liquid, but 

 at a little distance from it ; and consequently, as the contraction proceeds, 

 it will be at this spot that the union of the sides of the film and the separa- 

 tion will take place. This arises from the fact that this is the weakest 

 point of tension between the ring and the liquid, and therefore the one in 

 which circumferential contraction can take place with the greatest ease and 

 effect. Thus we see that the tubular film which we have raised really be- 

 comes constricted into two portions, — an upper portion, which immediately 

 contracts into a plane surface upon the ring, and a smaller and lower 

 portion, which, in consequence of including air, becomes a hemisphere and 

 remains attached to the surface of the solution. If, having formed such a 

 film upon a ring or pipe-bowl, we proceed to blow down upon it, we 

 distend it into a sphere ; but it is obvious that until the sphere is detached 



tion beyond the line or boundary of actual contact, All capillary phenomena may 

 be regarded as due to this* progressive action of cohesion operating at one and the 

 same time in diverse directions. 



