1871.] Blood-corpuscles through the Walls of the Vessels. 561 



laws as the soap-splieres, and we have only to infer that they hear the 

 same relation to the capillary walls as these spheres and films bear to each 

 other. The margin of speculation is therefore small. 



In the case of the corpuscles this relation is of course only seen under 

 abnormal conditions, simply because it is a physical law which in the 

 normal working of the animal economy required to be antagonized. 



It must also be observed that it is only under certain conditions that the 

 soap-spheres attract each other, or are attracted by rigid surfaces or plastic 

 films. This occurs only when free liquid is cohering to their surfaces. 

 If before bringing them into contact we allow the soap-film and sphere to 

 become moderately dry, they will not attract each other, but the former 

 will support the latter as a perfect sphere instead of drawing it down by 

 progressive cohesion and arranging it halfway through itself. 



Just so with the corpuscles ; they do not unite either with each other or 

 with the capillary wall, unless their normal osmotic relations are disturbed, 

 the exosmotic current setting in excessively when their external surfaces 

 become coated with content-matter, and they become instantly attractive 

 of each other, of the capillary wall or glass slide, as the case may be. 



In the paper before referred to, " On the Laws concerned in the Aggre- 

 gation of the Blood- corpuscles," I have given numerous examples of the 

 operation of progressive cohesive attraction ; but in this place I wish to call 

 attention to the demonstration there given of its relation to plane sur- 

 faces. 



Taking this experiment as a starting-point, we will extend the consi- 

 deration to surfaces of a different character. In the first place, we find 

 that this law continues to operate with great facility in connexion with 

 surfaces curved in one direction only, whether the surface used be convex 

 or concave ; in both cases the film of paper or collodion applies itself evenly 

 to the surface in the gradual progressive manner before explained. 



If, however, for surfaces curved in one direction only, we substitute such 

 as are curved in all directions (for example, the outer or convex, or the inner 

 or concave surfaces of a hollow sphere), we find ourselves confronted with 

 a new set of difficulties, out of which we may evolve the statement that, 

 for any film to apply itself evenly and regularly to either the convex or 

 concave surface of a sphere under the influence of progressive attraction, 

 it is necessary that the film should be, in several particulars of its consti- 

 tution, very different from the class of films by means of which we have 

 been able to illustrate the three preceding experiments. 



If, by way of illustration, we apply a film of wetted collodion or fine 

 cambric paper to the sphere, so that one point of the convexity of the 

 latter may come in contact with the centre of the film, the attraction will 

 only succeed in pulling it dow r n to the surface of the sphere at certain 

 points ; the intermediate puckered parts are not in contact, and can by no 

 possibility become applied. From this we see that for the film to be laid 

 down evenly, it would be necessary that it should contract in certain parts, 



vol. xix. 2 x 



