ARTERIES AND CAPILLARIES. 143 



larger ones, although even when full of blood they only 

 appear to the naked eye like red filaments, pass gradually 

 into smaller ones, and with a power magnifying three 

 hundred diameters, we see them breaking up into 

 branches, into which, even when they are very small, the 

 three coats are at first continued. It is only in the 

 smallest branches that the muscular coat finally disap- 

 pears, the intervals between the individual transverse 

 fibres becoming wider and wider, and the internal coat 

 (the nuclei of which lie in a longitudinal direction and 

 cross those of the middle coat at right angles (Fig. 26, 

 D, E,)), at the same time appearing more and more 

 distinctly through it. The external coat also may be 

 followed for a short distance farther (being in many 

 places, as in the brain, rendered more evident by the in- 

 terspersion of pigment or fat, Fig. 26, D, E\ till at last 

 it also becomes lost to view, and only a simple capillary 

 remains (Fig. 3, c). The general supposition, therefore, 

 is that the proper capillary membrane most nearly cor- 

 responds to the internal coat of the larger vessels, and it 

 is usually considered that the more complete a vessel 

 becomes, the greater is the number of the coats which 

 develop themselves around it. The real developmental 

 relations which these parts bear to one another have, 

 however, been by no means accurately determined. 



Within the true capillary region there is nothing 

 further worthy of notice in the vessels than the nuclei I 

 have previously mentioned, which correspond to the 

 longitudinal axis of the vessel, and are so imbedded in 

 its membrane, that it is impossible to discover any traces 

 of a surrounding cell- wall. The capillary membrane is 

 seen to be quite uniform, absolutely homogeneous and 

 continuous (Fig. 3, c). Whilst even as lately as twenty 

 years ago, it was a matter of discussion whether there 

 did not exist vessels which were destitute of true walls, 



